mirror of
https://github.com/NVIDIA/TensorRT-LLM.git
synced 2026-01-14 06:27:45 +08:00
12549 lines
2.7 MiB
12549 lines
2.7 MiB
|
||
|
||
<!DOCTYPE html>
|
||
<html class="writer-html5" lang="en" data-content_root="../">
|
||
<head>
|
||
<meta charset="utf-8" /><meta name="viewport" content="width=device-width, initial-scale=1" />
|
||
|
||
<meta name="viewport" content="width=device-width, initial-scale=1.0" />
|
||
<title>Runtime — tensorrt_llm documentation</title>
|
||
<link rel="stylesheet" type="text/css" href="../_static/pygments.css?v=80d5e7a1" />
|
||
<link rel="stylesheet" type="text/css" href="../_static/css/theme.css?v=e59714d7" />
|
||
<link rel="stylesheet" type="text/css" href="../_static/copybutton.css?v=76b2166b" />
|
||
|
||
|
||
<script src="../_static/jquery.js?v=5d32c60e"></script>
|
||
<script src="../_static/_sphinx_javascript_frameworks_compat.js?v=2cd50e6c"></script>
|
||
<script src="../_static/documentation_options.js?v=5929fcd5"></script>
|
||
<script src="../_static/doctools.js?v=9bcbadda"></script>
|
||
<script src="../_static/sphinx_highlight.js?v=dc90522c"></script>
|
||
<script src="../_static/clipboard.min.js?v=a7894cd8"></script>
|
||
<script src="../_static/copybutton.js?v=65e89d2a"></script>
|
||
<script src="../_static/js/theme.js"></script>
|
||
<link rel="index" title="Index" href="../genindex.html" />
|
||
<link rel="search" title="Search" href="../search.html" />
|
||
<link rel="next" title="trtllm-build" href="../commands/trtllm-build.html" />
|
||
<link rel="prev" title="Executor" href="executor.html" />
|
||
</head>
|
||
|
||
<body class="wy-body-for-nav">
|
||
<div class="wy-grid-for-nav">
|
||
<nav data-toggle="wy-nav-shift" class="wy-nav-side">
|
||
<div class="wy-side-scroll">
|
||
<div class="wy-side-nav-search" >
|
||
|
||
|
||
|
||
<a href="../index.html" class="icon icon-home">
|
||
tensorrt_llm
|
||
</a>
|
||
<div role="search">
|
||
<form id="rtd-search-form" class="wy-form" action="../search.html" method="get">
|
||
<input type="text" name="q" placeholder="Search docs" aria-label="Search docs" />
|
||
<input type="hidden" name="check_keywords" value="yes" />
|
||
<input type="hidden" name="area" value="default" />
|
||
</form>
|
||
</div>
|
||
</div><div class="wy-menu wy-menu-vertical" data-spy="affix" role="navigation" aria-label="Navigation menu">
|
||
<p class="caption" role="heading"><span class="caption-text">Getting Started</span></p>
|
||
<ul>
|
||
<li class="toctree-l1"><a class="reference internal" href="../overview.html">Overview</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../quick-start-guide.html">Quick Start Guide</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../key-features.html">Key Features</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../torch.html">PyTorch Backend</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../release-notes.html">Release Notes</a></li>
|
||
</ul>
|
||
<p class="caption" role="heading"><span class="caption-text">Installation</span></p>
|
||
<ul>
|
||
<li class="toctree-l1"><a class="reference internal" href="../installation/linux.html">Installing on Linux</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../installation/build-from-source-linux.html">Building from Source Code on Linux</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../installation/windows.html">Installing on Windows</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../installation/build-from-source-windows.html">Building from Source Code on Windows</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../installation/grace-hopper.html">Installing on Grace Hopper</a></li>
|
||
</ul>
|
||
<p class="caption" role="heading"><span class="caption-text">LLM API</span></p>
|
||
<ul>
|
||
<li class="toctree-l1"><a class="reference internal" href="../llm-api/index.html">API Introduction</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../llm-api/reference.html">API Reference</a></li>
|
||
</ul>
|
||
<p class="caption" role="heading"><span class="caption-text">LLM API Examples</span></p>
|
||
<ul>
|
||
<li class="toctree-l1"><a class="reference internal" href="../llm-api-examples/index.html">LLM Examples Introduction</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../llm-api-examples/customization.html">Common Customizations</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../llm-api-examples/llm_api_examples.html">Examples</a></li>
|
||
</ul>
|
||
<p class="caption" role="heading"><span class="caption-text">Model Definition API</span></p>
|
||
<ul>
|
||
<li class="toctree-l1"><a class="reference internal" href="../python-api/tensorrt_llm.layers.html">Layers</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../python-api/tensorrt_llm.functional.html">Functionals</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../python-api/tensorrt_llm.models.html">Models</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../python-api/tensorrt_llm.plugin.html">Plugin</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../python-api/tensorrt_llm.quantization.html">Quantization</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../python-api/tensorrt_llm.runtime.html">Runtime</a></li>
|
||
</ul>
|
||
<p class="caption" role="heading"><span class="caption-text">C++ API</span></p>
|
||
<ul class="current">
|
||
<li class="toctree-l1"><a class="reference internal" href="executor.html">Executor</a></li>
|
||
<li class="toctree-l1 current"><a class="current reference internal" href="#">Runtime</a><ul>
|
||
<li class="toctree-l2"><a class="reference internal" href="#buffermanager-h">bufferManager.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv412tensorrt_llm"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtimeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#common-h">common.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SizeType32</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType64E"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SizeType64</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11TokenIdTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::TokenIdType</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14LoraTaskIdTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraTaskIdType</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16TokenExtraIdTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::TokenExtraIdType</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16VecTokenExtraIdsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::VecTokenExtraIds</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15VecUniqueTokensE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::VecUniqueTokens</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime12StringPtrMapE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::StringPtrMap</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11RequestTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RequestType</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11RequestType8kCONTEXTE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RequestType::kCONTEXT</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11RequestType11kGENERATIONE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RequestType::kGENERATION</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11UniqueTokenE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::UniqueToken</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11UniqueTokeneqERK11UniqueToken"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::UniqueToken::operator==()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11UniqueToken7tokenIdE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::UniqueToken::tokenId</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11UniqueToken12tokenExtraIdE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::UniqueToken::tokenExtraId</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#cudaevent-h">cudaEvent.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEventE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::CudaEvent</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEvent7pointerE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::CudaEvent::pointer</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEvent9CudaEventEj"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::CudaEvent::CudaEvent()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEvent9CudaEventE7pointerb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::CudaEvent::CudaEvent()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime9CudaEvent3getEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::CudaEvent::get()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime9CudaEvent11synchronizeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::CudaEvent::synchronize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEvent12element_typeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::CudaEvent::element_type</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEvent8EventPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::CudaEvent::EventPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEvent6mEventE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::CudaEvent::mEvent</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEvent7DeleterE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::CudaEvent::Deleter</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#cudastream-h">cudaStream.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10CudaStreamE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::CudaStream</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10CudaStream10CudaStreamEji"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::CudaStream::CudaStream()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10CudaStream10CudaStreamE12cudaStream_tib"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::CudaStream::CudaStream()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10CudaStream10CudaStreamE12cudaStream_t"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::CudaStream::CudaStream()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10CudaStream9getDeviceEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::CudaStream::getDevice()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10CudaStream3getEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::CudaStream::get()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10CudaStream11synchronizeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::CudaStream::synchronize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10CudaStream6recordEN9CudaEvent7pointerE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::CudaStream::record()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10CudaStream6recordERK9CudaEvent"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::CudaStream::record()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10CudaStream4waitEN9CudaEvent7pointerE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::CudaStream::wait()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10CudaStream4waitERK9CudaEvent"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::CudaStream::wait()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10CudaStream9StreamPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::CudaStream::StreamPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10CudaStream7mStreamE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::CudaStream::mStream</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10CudaStream7mDeviceE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::CudaStream::mDevice</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10CudaStream7DeleterE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::CudaStream::Deleter</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#decodinginput-h">decodingInput.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInputE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::TensorConstPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput9TensorPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::TensorPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput13DecodingInputE10SizeType3210SizeType3210SizeType3210SizeType3214TensorConstPtr9TensorPtr14TensorConstPtr"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::DecodingInput()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput4stepE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::step</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput9maxLengthE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::maxLength</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput18maxAttentionWindowE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::maxAttentionWindow</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput15sinkTokenLengthE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::sinkTokenLength</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput9batchSizeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::batchSize</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput15maxStopWordsLenE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::maxStopWordsLen</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14maxBadWordsLenE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::maxBadWordsLen</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput6logitsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::logits</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput6endIdsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::endIds</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput10batchSlotsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::batchSlots</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput13finishReasonsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::finishReasons</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput19sequenceLimitLengthE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::sequenceLimitLength</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput13embeddingBiasE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::embeddingBias</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput7lengthsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::lengths</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput13badWordsListsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::badWordsLists</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput12badWordsPtrsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::badWordsPtrs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput12badWordsLensE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::badWordsLens</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14stopWordsListsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::stopWordsLists</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput13stopWordsPtrsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::stopWordsPtrs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput13stopWordsLensE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::stopWordsLens</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput17noRepeatNgramSizeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::noRepeatNgramSize</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput16cacheIndirectionE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::cacheIndirection</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput12medusaInputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::medusaInputs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25explicitDraftTokensInputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::explicitDraftTokensInputs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput15lookaheadInputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::lookaheadInputs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25externalDraftTokensInputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::externalDraftTokensInputs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput11eagleInputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::eagleInputs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::EagleInputs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::ExplicitDraftTokensInputs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::ExternalDraftTokensInputs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput15LookaheadInputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::LookaheadInputs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput12MedusaInputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingInput::MedusaInputs</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#decodingoutput-h">decodingOutput.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm13batch_managerE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::batch_manager</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutputE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingOutput</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingOutput::TensorPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14DecodingOutputE9TensorPtr9TensorPtr"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingOutput::DecodingOutput()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput3idsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingOutput::ids</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput11gatheredIdsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingOutput::gatheredIds</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14newTokensStepsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingOutput::newTokensSteps</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9newTokensE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingOutput::newTokens</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput12newTokensVecE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingOutput::newTokensVec</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput13finishReasonsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingOutput::finishReasons</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput11finishedSumE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingOutput::finishedSum</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput8logProbsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingOutput::logProbs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput11cumLogProbsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingOutput::cumLogProbs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9parentIdsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingOutput::parentIds</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput7lengthsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingOutput::lengths</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput16cacheIndirectionE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingOutput::cacheIndirection</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput13logProbsTiledE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingOutput::logProbsTiled</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14beamHypothesesE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingOutput::beamHypotheses</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput26speculativeDecodingOutputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingOutput::speculativeDecodingOutputs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput26explicitDraftTokensBuffersE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingOutput::explicitDraftTokensBuffers</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput16lookaheadOutputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingOutput::lookaheadOutputs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput12eagleBuffersE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingOutput::eagleBuffers</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput17kNegativeInfinityE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingOutput::kNegativeInfinity</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypothesesE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingOutput::BeamHypotheses</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DecodingOutput::SpeculativeDecodingOutputs</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#generationinput-h">generationInput.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15GenerationInputE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenerationInput</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15GenerationInput4BaseE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenerationInput::Base</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15GenerationInput9TensorPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenerationInput::TensorPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15GenerationInput15GenerationInputEK10SizeType32K10SizeType329TensorPtr9TensorPtrb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenerationInput::GenerationInput()</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4I00EN12tensorrt_llm7runtime22GenericGenerationInputE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericGenerationInput</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput9TensorPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericGenerationInput::TensorPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput22GenericGenerationInputEK10SizeType32K10SizeType329TensorPtr9TensorPtrb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericGenerationInput::GenericGenerationInput()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput5endIdE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericGenerationInput::endId</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput5padIdE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericGenerationInput::padId</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput3idsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericGenerationInput::ids</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput7lengthsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericGenerationInput::lengths</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput6packedE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericGenerationInput::packed</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput13embeddingBiasE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericGenerationInput::embeddingBias</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput12badWordsListE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericGenerationInput::badWordsList</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput13stopWordsListE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericGenerationInput::stopWordsList</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput12maxNewTokensE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericGenerationInput::maxNewTokens</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput18promptTuningParamsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericGenerationInput::promptTuningParams</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#generationoutput-h">generationOutput.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16GenerationOutputE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenerationOutput</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16GenerationOutput4BaseE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenerationOutput::Base</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16GenerationOutput9TensorPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenerationOutput::TensorPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16GenerationOutput16GenerationOutputE9TensorPtr9TensorPtr"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenerationOutput::GenerationOutput()</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime23GenericGenerationOutputE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericGenerationOutput</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput9TensorPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericGenerationOutput::TensorPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput8CallbackE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericGenerationOutput::Callback</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput23GenericGenerationOutputE9TensorPtr9TensorPtr"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericGenerationOutput::GenericGenerationOutput()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput3idsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericGenerationOutput::ids</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput7lengthsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericGenerationOutput::lengths</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput11cumLogProbsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericGenerationOutput::cumLogProbs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput8logProbsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericGenerationOutput::logProbs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput13contextLogitsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericGenerationOutput::contextLogits</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput16generationLogitsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericGenerationOutput::generationLogits</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput16onTokenGeneratedE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericGenerationOutput::onTokenGenerated</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#gptjsonconfig-h">gptJsonConfig.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptJsonConfig</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig13GptJsonConfigENSt6stringENSt6stringENSt6stringE10SizeType3210SizeType3210SizeType3210SizeType3211ModelConfigNSt8optionalI15RuntimeDefaultsEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptJsonConfig::GptJsonConfig()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig14getModelConfigEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptJsonConfig::getModelConfig()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig21getModelConfigMutableEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptJsonConfig::getModelConfigMutable()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig7getNameEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptJsonConfig::getName()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig10getVersionEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptJsonConfig::getVersion()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig12getPrecisionEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptJsonConfig::getPrecision()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig20getTensorParallelismEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptJsonConfig::getTensorParallelism()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig22getPipelineParallelismEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptJsonConfig::getPipelineParallelism()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig21getContextParallelismEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptJsonConfig::getContextParallelism()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig14getGpusPerNodeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptJsonConfig::getGpusPerNode()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig12getWorldSizeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptJsonConfig::getWorldSize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig18getRuntimeDefaultsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptJsonConfig::getRuntimeDefaults()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig14engineFilenameERK11WorldConfigRKNSt6stringE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptJsonConfig::engineFilename()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig14engineFilenameERK11WorldConfig"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptJsonConfig::engineFilename()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig5parseERKNSt6stringE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptJsonConfig::parse()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig5parseERNSt7istreamE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptJsonConfig::parse()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig5parseERKNSt10filesystem4pathE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptJsonConfig::parse()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig5mNameE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptJsonConfig::mName</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig8mVersionE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptJsonConfig::mVersion</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig10mPrecisionE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptJsonConfig::mPrecision</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig18mTensorParallelismE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptJsonConfig::mTensorParallelism</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig20mPipelineParallelismE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptJsonConfig::mPipelineParallelism</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig19mContextParallelismE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptJsonConfig::mContextParallelism</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig12mGpusPerNodeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptJsonConfig::mGpusPerNode</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig12mModelConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptJsonConfig::mModelConfig</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig16mRuntimeDefaultsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptJsonConfig::mRuntimeDefaults</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#itensor-h">iTensor.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv48nvinfer1"><code class="docutils literal notranslate"><span class="pre">nvinfer1</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtimelsERNSt7ostreamERKN7ITensor5ShapeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::operator<<()</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtimelsERNSt7ostreamERK7ITensor"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::operator<<()</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEPK1TRKN7ITensor14SharedConstPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::bufferCastOrNull()</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEP1TRKN7ITensor9SharedPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::bufferCastOrNull()</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEP1TRKNSt8optionalIN7ITensor9SharedPtrEEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::bufferCastOrNull()</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEPK1TRKNSt8optionalIN7ITensor14SharedConstPtrEEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::bufferCastOrNull()</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9UniquePtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::UniquePtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::SharedPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor14UniqueConstPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::UniqueConstPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor14SharedConstPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::SharedConstPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::Shape</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9DimType64E"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::DimType64</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9TensorMapE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::TensorMap</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorD0Ev"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::~ITensor()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime7ITensor8getShapeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::getShape()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I_10SizeType32ENK12tensorrt_llm7runtime7ITensor12getDimensionE9DimType64v"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::getDimension()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor7reshapeERK5Shape"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::reshape()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor6resizeENSt6size_tE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::resize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor7ITensorERK7ITensor"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::ITensor()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensoraSERK7ITensor"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::operator=()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor7squeezeE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::squeeze()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9unsqueezeE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::unsqueeze()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime7ITensor11shapeEqualsERK5Shape"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::shapeEquals()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime7ITensor11shapeEqualsERKNSt16initializer_listI10SizeType32EE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::shapeEquals()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0ENK12tensorrt_llm7runtime7ITensor11shapeEqualsEbPK1T10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::shapeEquals()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor6volumeERK5Shape"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::volume()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor17volumeNonNegativeERK5Shape"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::volumeNonNegative()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor7stridesERK5Shape"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::strides()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor7squeezeERK5Shape10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::squeeze()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9unsqueezeERK5Shape10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::unsqueeze()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrNSt6size_tENSt6size_tE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::slice()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrNSt6size_tENSt6size_tE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::slice()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrNSt6size_tE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::slice()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrNSt6size_tE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::slice()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrRK5Shape9DimType64"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::slice()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrRKNSt16initializer_listI9DimType64EE9DimType64"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::slice()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrRK5ShapeNSt6size_tE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::slice()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrRKNSt16initializer_listI9DimType64EENSt6size_tE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::slice()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrRK5Shape"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::slice()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrRKNSt16initializer_listI9DimType64EE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::slice()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrRK5Shape"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::slice()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrRKNSt16initializer_listI9DimType64EE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::slice()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor2atE9SharedPtrRK5Shape"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::at()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor2atE9SharedPtrRKNSt16initializer_listI9DimType64EE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::at()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor2atE14UniqueConstPtrRR9TConstPtrRK5Shape"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::at()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor2atEN7ITensor14UniqueConstPtrERR9TConstPtrRKNSt16initializer_listI9DimType64EE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::at()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor4viewEN7IBuffer9SharedPtrERK5Shape"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::view()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor4viewE14UniqueConstPtrRR9TConstPtrRK5Shape"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::view()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor4viewE9SharedPtr"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::view()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor8flattenNE9SharedPtrNSt7int64_tE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::flattenN()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor4wrapEPvN8nvinfer18DataTypeERK5ShapeNSt6size_tE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::wrap()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor4wrapEPvN8nvinfer18DataTypeERK5Shape"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::wrap()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime7ITensor4wrapE9UniquePtrP1TRK5ShapeNSt6size_tE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::wrap()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime7ITensor4wrapE9UniquePtrP1TRK5Shape"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::wrap()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime7ITensor4wrapE9UniquePtrRNSt6vectorI1TEERK5Shape"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::wrap()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9makeShapeERKNSt16initializer_listI9DimType64EE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::makeShape()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor8toStringERK5Shape"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::toString()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor11shapeEqualsERK5ShapeRK5Shape"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::shapeEquals()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime7ITensor11shapeEqualsEbRK5ShapePK1T10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::shapeEquals()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor7ITensorEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::ITensor()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor8castSizeE6size_t"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ITensor::castSize()</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#ipcutils-h">ipcUtils.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime20lamportInitializeAllEPvPvPv6size_t"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::lamportInitializeAll()</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16AllReduceBuffersE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::AllReduceBuffers</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16AllReduceBuffers9TensorPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::AllReduceBuffers::TensorPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16AllReduceBuffers16AllReduceBuffersE10SizeType3210SizeType3210SizeType3210SizeType32RK13BufferManagerRK11WorldConfigKb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::AllReduceBuffers::AllReduceBuffers()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16AllReduceBuffers18mAllReduceCommPtrsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::AllReduceBuffers::mAllReduceCommPtrs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16AllReduceBuffers17mIpcMemoryHandlesE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::AllReduceBuffers::mIpcMemoryHandles</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemoryE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IpcMemory</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory9BufferPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IpcMemory::BufferPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory9IpcMemoryENSt6size_tERK13BufferManagerRK11WorldConfigb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IpcMemory::IpcMemory()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemoryD0Ev"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IpcMemory::~IpcMemory()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory9IpcMemoryERK9IpcMemory"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IpcMemory::IpcMemory()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemoryaSERK9IpcMemory"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IpcMemory::operator=()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory9IpcMemoryERR9IpcMemory"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IpcMemory::IpcMemory()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemoryaSERR9IpcMemory"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IpcMemory::operator=()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime9IpcMemory11getCommPtrsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IpcMemory::getCommPtrs()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory10FLAGS_SIZEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IpcMemory::FLAGS_SIZE</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory17allocateIpcMemoryENSt6size_tERK13BufferManagerRK11WorldConfig"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IpcMemory::allocateIpcMemory()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory16destroyIpcMemoryEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IpcMemory::destroyIpcMemory()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory7mTpRankE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IpcMemory::mTpRank</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory9mCommPtrsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IpcMemory::mCommPtrs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory7mBufferE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IpcMemory::mBuffer</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory8mOpenIpcE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IpcMemory::mOpenIpc</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#loracache-h">loraCache.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9to_stringERKN9LoraCache21TaskLayerModuleConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::to_string()</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtimelsERNSt7ostreamERKN9LoraCache21TaskLayerModuleConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::operator<<()</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCacheE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TensorPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::TensorPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache10TaskIdTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::TaskIdType</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache28TaskLayerModuleConfigListPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::TaskLayerModuleConfigListPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9LoraCacheERK26LoraCachePageManagerConfigRK11ModelConfigRK11WorldConfigRK13BufferManager"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::LoraCache()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache3putE10TaskIdType9TensorPtr9TensorPtrb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::put()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache11loadWeightsE10TaskIdType9TensorPtr9TensorPtr"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::loadWeights()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime9LoraCache8isLoadedE10TaskIdType"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::isLoaded()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime9LoraCache6isDoneE10TaskIdType"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::isDone()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime9LoraCache3hasE10TaskIdType"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::has()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache3getE10TaskIdType"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::get()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache4bumpE10TaskIdType"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::bump()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache12markTaskDoneE10TaskIdType"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::markTaskDone()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache11markAllDoneEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::markAllDone()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime9LoraCache17determineNumPagesE10TaskIdType"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::determineNumPages()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime9LoraCache17determineNumPagesE9TensorPtr"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::determineNumPages()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime9LoraCache4fitsE9TensorPtr"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::fits()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache8copyTaskE10TaskIdTypeR9LoraCacheb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::copyTask()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime9LoraCache11getNumPagesEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::getNumPages()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime9LoraCache10getPagePtrE6size_t"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::getPagePtr()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache11copyToPagesE9TensorPtr9TensorPtrRK11ModelConfigRK11WorldConfigNSt13unordered_mapI10SizeType3210LoraModuleEERK13BufferManagerRKNSt6vectorI9TensorPtrEERKNSt6vectorINSt6size_tEEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::copyToPages()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache17splitTransposeCpuER7ITensorRK7ITensor10SizeType3210SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::splitTransposeCpu()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache11ValueStatusE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::ValueStatus</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache12TaskValuePtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::TaskValuePtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache11loadWeightsER9TaskValue9TensorPtr9TensorPtr"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::loadWeights()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache18bumpTaskInProgressE10TaskIdType"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::bumpTaskInProgress()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime9LoraCache9getStatusE10TaskIdType"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::getStatus()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache19claimPagesWithEvictE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::claimPagesWithEvict()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache16copyTaskMapPagesER9TaskValueRK9TaskValueRKNSt6vectorI6size_tEERK9LoraCache"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::copyTaskMapPages()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache18mPageManagerConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::mPageManagerConfig</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache12mModelConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::mModelConfig</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache12mWorldConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::mWorldConfig</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache11mPagesMutexE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::mPagesMutex</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache17mCachePageManagerE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::mCachePageManager</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache11mCacheMutexE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::mCacheMutex</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9mCacheMapE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::mCacheMap</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache16mInProgressTasksE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::mInProgressTasks</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache10mDoneTasksE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::mDoneTasks</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache21mDeviceBufferManagersE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::mDeviceBufferManagers</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache14mBufferManagerE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::mBufferManager</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache17mModuleIdToModuleE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::mModuleIdToModule</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime9LoraCache22splitTransposeCpuInnerEvR7ITensorRK7ITensor10SizeType3210SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::splitTransposeCpuInner()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::TaskLayerModuleConfig</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValueE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCache::TaskValue</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime22LoraCacheFullExceptionE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCacheFullException</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime22LoraCacheFullException22LoraCacheFullExceptionERKNSt6stringE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCacheFullException::LoraCacheFullException()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime22LoraCacheFullExceptionD0Ev"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCacheFullException::~LoraCacheFullException()</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime20LoraCachePageManagerE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManager</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager9TensorPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManager::TensorPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager20LoraCachePageManagerERK26LoraCachePageManagerConfigRK13BufferManager"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManager::LoraCachePageManager()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager10claimPagesE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManager::claimPages()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime20LoraCachePageManager17numAvailablePagesEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManager::numAvailablePages()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager12releasePagesERKNSt6vectorINSt6size_tEEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManager::releasePages()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime20LoraCachePageManager8blockPtrE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManager::blockPtr()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime20LoraCachePageManager7pagePtrENSt6size_tE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManager::pagePtr()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager14mutablePagePtrENSt6size_tE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManager::mutablePagePtr()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager10initializeERK13BufferManager"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManager::initialize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager11mPageBlocksE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManager::mPageBlocks</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager12mFreePageIdsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManager::mFreePageIds</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager11mIsPageFreeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManager::mIsPageFree</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager7mConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManager::mConfig</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime21LoraExpectedExceptionE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraExpectedException</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime21LoraExpectedException21LoraExpectedExceptionERKNSt6stringE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraExpectedException::LoraExpectedException()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime21LoraExpectedExceptionD0Ev"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraExpectedException::~LoraExpectedException()</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#loracachepagemanagerconfig-h">loraCachePageManagerConfig.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtimelsERNSt7ostreamERK26LoraCachePageManagerConfig"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::operator<<()</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9to_stringERK26LoraCachePageManagerConfig"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::to_string()</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManagerConfig</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig26LoraCachePageManagerConfigEN7runtime10MemoryTypeEN8nvinfer18DataTypeE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManagerConfig::LoraCachePageManagerConfig()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime26LoraCachePageManagerConfig13getMemoryTypeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManagerConfig::getMemoryType()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig13setMemoryTypeERKN7runtime10MemoryTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManagerConfig::setMemoryType()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime26LoraCachePageManagerConfig11getDataTypeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManagerConfig::getDataType()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig11setDataTypeERKN8nvinfer18DataTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManagerConfig::setDataType()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime26LoraCachePageManagerConfig16getTotalNumPagesEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManagerConfig::getTotalNumPages()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig15setTotalNumPageERK10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManagerConfig::setTotalNumPage()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime26LoraCachePageManagerConfig19getMaxPagesPerBlockEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManagerConfig::getMaxPagesPerBlock()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig19setMaxPagesPerBlockERK10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManagerConfig::setMaxPagesPerBlock()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime26LoraCachePageManagerConfig15getSlotsPerPageEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManagerConfig::getSlotsPerPage()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig15setSlotsPerPageERK10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManagerConfig::setSlotsPerPage()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime26LoraCachePageManagerConfig12getPageWidthEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManagerConfig::getPageWidth()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig12setPageWidthERK10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManagerConfig::setPageWidth()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime26LoraCachePageManagerConfig13getInitToZeroEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManagerConfig::getInitToZero()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig13setInitToZeroEb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManagerConfig::setInitToZero()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime26LoraCachePageManagerConfig17getNumCopyStreamsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManagerConfig::getNumCopyStreams()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig17setNumCopyStreamsE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManagerConfig::setNumCopyStreams()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig11mMemoryTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManagerConfig::mMemoryType</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig9mDataTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManagerConfig::mDataType</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig14mTotalNumPagesE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManagerConfig::mTotalNumPages</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig17mMaxPagesPerBlockE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManagerConfig::mMaxPagesPerBlock</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig13mSlotsPerPageE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManagerConfig::mSlotsPerPage</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig10mPageWidthE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManagerConfig::mPageWidth</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig15mNumCopyStreamsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManagerConfig::mNumCopyStreams</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig11mInitToZeroE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraCachePageManagerConfig::mInitToZero</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#loramodule-h">loraModule.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtimelsERNSt7ostreamERK10LoraModule"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::operator<<()</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModuleE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::ModuleType</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule9TensorPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::TensorPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10LoraModuleERK10ModuleType10SizeType3210SizeType32bb10SizeType3210SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::LoraModule()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10LoraModuleEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::LoraModule()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10LoraModuleERK10LoraModule"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::LoraModule()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModuleaSERK10LoraModule"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::operator=()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule18flattenedInOutSizeE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::flattenedInOutSize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule6inSizeE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::inSize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule7outSizeE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::outSize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule11localInSizeE10SizeType3210SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::localInSize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule12localOutSizeE10SizeType3210SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::localOutSize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule10localInDimE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::localInDim()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule11localOutDimE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::localOutDim()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule18localInAdapterSizeE10SizeType3210SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::localInAdapterSize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule19localOutAdapterSizeE10SizeType3210SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::localOutAdapterSize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule14localInOutSizeE10SizeType3210SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::localInOutSize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule5valueEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::value()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule4nameEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::name()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule5inDimEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::inDim()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule6outDimEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::outDim()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule10inDimFirstEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::inDimFirst()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule11outDimFirstEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::outDimFirst()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule12inTpSplitDimEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::inTpSplitDim()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule13outTpSplitDimEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::outTpSplitDim()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule17createLoraModulesERKNSt6vectorINSt6stringEEE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType3210SizeType3210SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::createLoraModules()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule12toModuleTypeERKNSt11string_viewE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::toModuleType()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule12toModuleNameE10ModuleType"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::toModuleName()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule12toModuleNameE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::toModuleName()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule5mTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::mType</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule6mInDimE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::mInDim</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule7mOutDimE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::mOutDim</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule11mInDimFirstE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::mInDimFirst</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule12mOutDimFirstE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::mOutDimFirst</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule13mInTpSplitDimE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::mInTpSplitDim</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule14mOutTpSplitDimE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LoraModule::mOutTpSplitDim</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#medusamodule-h">medusaModule.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12MedusaModuleE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MedusaModule</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12MedusaModule9TensorPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MedusaModule::TensorPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12MedusaModule13MedusaChoicesE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MedusaModule::MedusaChoices</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12MedusaModule12MedusaModuleE10SizeType3210SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MedusaModule::MedusaModule()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12MedusaModule12MedusaModuleEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MedusaModule::MedusaModule()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime12MedusaModule16getMedusaChoicesEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MedusaModule::getMedusaChoices()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12MedusaModule21mDefaultMedusaChoicesE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MedusaModule::mDefaultMedusaChoices</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#memorycounters-h">memoryCounters.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCountersE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters10SizeType32E"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::SizeType32</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters8DiffTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::DiffType</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters14MemoryCountersEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::MemoryCounters()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime14MemoryCounters6getGpuEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::getGpu()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime14MemoryCounters6getCpuEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::getCpu()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime14MemoryCounters9getPinnedEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::getPinned()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime14MemoryCounters6getUVMEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::getUVM()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime14MemoryCounters13getPinnedPoolEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::getPinnedPool()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime14MemoryCounters10getGpuDiffEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::getGpuDiff()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime14MemoryCounters10getCpuDiffEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::getCpuDiff()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime14MemoryCounters13getPinnedDiffEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::getPinnedDiff()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime14MemoryCounters10getUVMDiffEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::getUVMDiff()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime14MemoryCounters17getPinnedPoolDiffEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::getPinnedPoolDiff()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I_10MemoryTypeEN12tensorrt_llm7runtime14MemoryCounters8allocateEv10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::allocate()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters8allocateE10MemoryType10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::allocate()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I_10MemoryTypeEN12tensorrt_llm7runtime14MemoryCounters10deallocateEv10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::deallocate()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters10deallocateE10MemoryType10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::deallocate()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime14MemoryCounters8toStringEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::toString()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters11getInstanceEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::getInstance()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters13bytesToStringE10SizeType32i"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::bytesToString()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters13bytesToStringE8DiffTypei"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::bytesToString()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters4mGpuE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::mGpu</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters4mCpuE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::mCpu</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters7mPinnedE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::mPinned</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters4mUVME"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::mUVM</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters11mPinnedPoolE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::mPinnedPool</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters8mGpuDiffE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::mGpuDiff</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters8mCpuDiffE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::mCpuDiff</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters11mPinnedDiffE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::mPinnedDiff</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters8mUVMDiffE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::mUVMDiff</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters15mPinnedPoolDiffE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryCounters::mPinnedPoolDiff</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#prompttuningparams-h">promptTuningParams.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime25GenericPromptTuningParamsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericPromptTuningParams</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams9TensorPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericPromptTuningParams::TensorPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams10SizeType32E"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericPromptTuningParams::SizeType32</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams25GenericPromptTuningParamsE9TensorPtr9TensorPtr9TensorPtr"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericPromptTuningParams::GenericPromptTuningParams()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams14embeddingTableE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericPromptTuningParams::embeddingTable</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams5tasksE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericPromptTuningParams::tasks</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams9vocabSizeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericPromptTuningParams::vocabSize</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams19promptTuningEnabledE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GenericPromptTuningParams::promptTuningEnabled</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18PromptTuningParamsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::PromptTuningParams</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18PromptTuningParams9TensorPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::PromptTuningParams::TensorPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18PromptTuningParams10SizeType32E"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::PromptTuningParams::SizeType32</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18PromptTuningParams18PromptTuningParamsE9TensorPtr9TensorPtr9TensorPtr"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::PromptTuningParams::PromptTuningParams()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18PromptTuningParams15fillTasksTensorE9TensorPtrK10SizeType32K10SizeType32RKNSt6vectorI10SizeType32EERKNSt6vectorI10SizeType32EERK13BufferManagerb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::PromptTuningParams::fillTasksTensor()</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#rawengine-h">rawEngine.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9RawEngineE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RawEngine</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine4TypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RawEngine::Type</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine9RawEngineENSt10filesystem4pathE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RawEngine::RawEngine()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine9RawEngineEPKvNSt6size_tE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RawEngine::RawEngine()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine9RawEngineEPKN8nvinfer111IHostMemoryE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RawEngine::RawEngine()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime9RawEngine7getTypeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RawEngine::getType()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime9RawEngine7getPathEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RawEngine::getPath()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime9RawEngine10getPathOptEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RawEngine::getPathOpt()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine7setPathENSt10filesystem4pathE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RawEngine::setPath()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime9RawEngine23getManagedWeightsMapOptEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RawEngine::getManagedWeightsMapOpt()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine20setManagedWeightsMapENSt3mapINSt6stringEN12tensorrt_llm8executor6TensorEEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RawEngine::setManagedWeightsMap()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime9RawEngine10getAddressEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RawEngine::getAddress()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime9RawEngine7getSizeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RawEngine::getSize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime9RawEngine13getHostMemoryEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RawEngine::getHostMemory()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine11mEngineAddrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RawEngine::mEngineAddr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine11mEngineSizeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RawEngine::mEngineSize</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine5mTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RawEngine::mType</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine11mEnginePathE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RawEngine::mEnginePath</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine13mEngineBufferE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RawEngine::mEngineBuffer</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine18mManagedWeightsMapE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RawEngine::mManagedWeightsMap</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#request-h">request.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::decoder_batch</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7RequestE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::decoder_batch::Request</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#runtimedefaults-h">runtimeDefaults.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15RuntimeDefaultsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RuntimeDefaults</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15RuntimeDefaults15RuntimeDefaultsENSt8optionalINSt6vectorI10SizeType32EEEENSt8optionalI10SizeType32EE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RuntimeDefaults::RuntimeDefaults()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15RuntimeDefaults15RuntimeDefaultsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RuntimeDefaults::RuntimeDefaults()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15RuntimeDefaults21maxAttentionWindowVecE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RuntimeDefaults::maxAttentionWindowVec</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15RuntimeDefaults15sinkTokenLengthE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::RuntimeDefaults::sinkTokenLength</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#speculativedecodingmode-h">speculativeDecodingMode.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingModeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode14UnderlyingTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::UnderlyingType</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode6isNoneEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::isNone()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode21isDraftTokensExternalEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::isDraftTokensExternal()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode8isMedusaEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::isMedusa()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode19isLookaheadDecodingEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::isLookaheadDecoding()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode21isExplicitDraftTokensEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::isExplicitDraftTokens()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode7isEagleEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::isEagle()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode18updatesPositionIdsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::updatesPositionIds()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode21requiresAttentionMaskEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::requiresAttentionMask()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode19predictsDraftTokensEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::predictsDraftTokens()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode18needsKVCacheRewindEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::needsKVCacheRewind()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode19variableDraftLengthEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::variableDraftLength()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode14hasDraftLogitsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::hasDraftLogits()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode20needsDecoderPrologueEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::needsDecoderPrologue()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingModeeqERK23SpeculativeDecodingMode"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::operator==()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode23SpeculativeDecodingModeE14UnderlyingType"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::SpeculativeDecodingMode()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode4NoneEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::None()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode19DraftTokensExternalEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::DraftTokensExternal()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode6MedusaEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::Medusa()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode17LookaheadDecodingEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::LookaheadDecoding()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode19ExplicitDraftTokensEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::ExplicitDraftTokens()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode5EagleEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::Eagle()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode9anyBitSetE14UnderlyingType"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::anyBitSet()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode9allBitSetE14UnderlyingType"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::allBitSet()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode6mStateE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::mState</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode5kNoneE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::kNone</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode20kDraftTokensExternalE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::kDraftTokensExternal</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode7kMedusaE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::kMedusa</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode18kLookaheadDecodingE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::kLookaheadDecoding</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode20kExplicitDraftTokensE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::kExplicitDraftTokens</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode6kEagleE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingMode::kEagle</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#speculativedecodingmodule-h">speculativeDecodingModule.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModuleE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingModule</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule25SpeculativeDecodingModuleE10SizeType3210SizeType3210SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingModule::SpeculativeDecodingModule()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule25SpeculativeDecodingModuleEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingModule::SpeculativeDecodingModule()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModuleD0Ev"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingModule::~SpeculativeDecodingModule()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule25SpeculativeDecodingModuleERK25SpeculativeDecodingModule"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingModule::SpeculativeDecodingModule()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModuleaSERK25SpeculativeDecodingModule"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingModule::operator=()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime25SpeculativeDecodingModule18getMaxDraftPathLenEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingModule::getMaxDraftPathLen()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime25SpeculativeDecodingModule13getMaxPathLenEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingModule::getMaxPathLen()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime25SpeculativeDecodingModule25getMaxDecodingDraftTokensEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingModule::getMaxDecodingDraftTokens()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime25SpeculativeDecodingModule20getMaxDecodingTokensEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingModule::getMaxDecodingTokens()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime25SpeculativeDecodingModule17getNumPackedMasksEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingModule::getNumPackedMasks()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime25SpeculativeDecodingModule14getMaxNumPathsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingModule::getMaxNumPaths()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule17setMaxDraftTokensE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingModule::setMaxDraftTokens()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule18setMaxDraftPathLenE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingModule::setMaxDraftPathLen()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule14setMaxNumPathsE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingModule::setMaxNumPaths()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule21computeNumPackedMasksEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingModule::computeNumPackedMasks()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule16mMaxDraftPathLenE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingModule::mMaxDraftPathLen</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule23mMaxDecodingDraftTokensE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingModule::mMaxDecodingDraftTokens</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule12mMaxNumPathsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingModule::mMaxNumPaths</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule18mMaxNumPackedMasksE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SpeculativeDecodingModule::mMaxNumPackedMasks</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#tllmlogger-h">tllmLogger.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10TllmLoggerE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::TllmLogger</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10TllmLogger3logE8SeverityPKN8nvinfer19AsciiCharE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::TllmLogger::log()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10TllmLogger8getLevelEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::TllmLogger::getLevel()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10TllmLogger8setLevelE8Severity"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::TllmLogger::setLevel()</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#worldconfig-h">worldConfig.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfig11WorldConfigE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType32RKNSt8optionalINSt6vectorI10SizeType32EEEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::WorldConfig()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig7getSizeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::getSize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig20getTensorParallelismEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::getTensorParallelism()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig16isTensorParallelEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::isTensorParallel()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig22getPipelineParallelismEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::getPipelineParallelism()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig18isPipelineParallelEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::isPipelineParallel()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig21getContextParallelismEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::getContextParallelism()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig17isContextParallelEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::isContextParallel()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig7getRankEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::getRank()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig14getGpusPerNodeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::getGpusPerNode()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig15getGpusPerGroupEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::getGpusPerGroup()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig9getDeviceEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::getDevice()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig11getDeviceOfE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::getDeviceOf()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig23getPipelineParallelRankEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::getPipelineParallelRank()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig21getTensorParallelRankEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::getTensorParallelRank()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig22getContextParallelRankEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::getContextParallelRank()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig12getLocalRankEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::getLocalRank()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig11getNodeRankEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::getNodeRank()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig13getNodeRankOfE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::getNodeRankOf()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig27isFirstPipelineParallelRankEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::isFirstPipelineParallelRank()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig26isLastPipelineParallelRankEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::isLastPipelineParallelRank()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig25isFirstTensorParallelRankEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::isFirstTensorParallelRank()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig26isFirstContextParallelRankEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::isFirstContextParallelRank()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig11getLastRankEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::getLastRank()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig24getPipelineParallelGroupEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::getPipelineParallelGroup()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig22getTensorParallelGroupEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::getTensorParallelGroup()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig23getContextParallelGroupEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::getContextParallelGroup()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig14validMpiConfigEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::validMpiConfig()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfig3mpiE10SizeType32NSt8optionalI10SizeType32EENSt8optionalI10SizeType32EENSt8optionalI10SizeType32EERKNSt8optionalINSt6vectorI10SizeType32EEEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::mpi()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfig19kDefaultGpusPerNodeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::kDefaultGpusPerNode</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfig18mTensorParallelismE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::mTensorParallelism</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfig20mPipelineParallelismE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::mPipelineParallelism</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfig19mContextParallelismE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::mContextParallelism</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfig5mRankE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::mRank</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfig12mGpusPerNodeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::mGpusPerNode</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfig10mDeviceIdsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::WorldConfig::mDeviceIds</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#eaglebuffers-h">eagleBuffers.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffersE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers13LlmRequestPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers::LlmRequestPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers13RequestVectorE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers::RequestVector</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers10SizeType32E"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers::SizeType32</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers7ITensorE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers::ITensor</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9BufferPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers::BufferPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers::TensorPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorMapE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers::TensorMap</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers12EagleBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerERKN7runtime11ModelConfigERKN7runtime11WorldConfigERKN8executor14DecodingConfigERKN7runtime11TllmRuntimeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers::EagleBuffers()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers7reshapeE10SizeType3210SizeType32RKN7runtime11ModelConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers::reshape()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime12EagleBuffers13setFromInputsERK13RequestVectorRK13RequestVectorRKN7runtime7ITensorERK7ITensorRKN12EagleBuffers6InputsERKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers::setFromInputs()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime12EagleBuffers18insertInputTensorsER9TensorMapR9TensorMapRKN7runtime11WorldConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers::insertInputTensors()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers12engineInputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers::engineInputs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers13engineOutputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers::engineOutputs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0ENK12tensorrt_llm7runtime12EagleBuffers13setFromInputsEvRK13RequestVectorRK13RequestVector10SizeType32RK7ITensorRKN12EagleBuffers6InputsERKN7runtime11EagleModuleERKN7runtime13BufferManagerE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers::setFromInputs()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers20scanTempStorageBytesE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers::scanTempStorageBytes</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers22reduceTempStorageBytesE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers::reduceTempStorageBytes</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers26mDefaultPosteriorThresholdE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers::mDefaultPosteriorThreshold</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers17mDoGreedySamplingE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers::mDoGreedySampling</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers21scanReduceTempStorageE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers::scanReduceTempStorage</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers23cumSumGenerationLengthsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers::cumSumGenerationLengths</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers19maxGenerationLengthE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers::maxGenerationLength</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers28chunkedContextNextTokensHostE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers::chunkedContextNextTokensHost</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers18greedySamplingHostE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers::greedySamplingHost</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers18posteriorAlphaHostE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers::posteriorAlphaHost</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers22posteriorThresholdHostE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers::posteriorThresholdHost</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers13EngineOutputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers::EngineOutputs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6InputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::EagleBuffers::Inputs</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#explicitdrafttokensbuffers-h">explicitDraftTokensBuffers.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffersE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ExplicitDraftTokensBuffers</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers10SizeType32E"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ExplicitDraftTokensBuffers::SizeType32</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers7ITensorE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ExplicitDraftTokensBuffers::ITensor</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9BufferPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ExplicitDraftTokensBuffers::BufferPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorMapE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorMap</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers26ExplicitDraftTokensBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerERKN7runtime11ModelConfigERKN7runtime11WorldConfigERKN8executor14DecodingConfigERKN7runtime11TllmRuntimeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ExplicitDraftTokensBuffers::ExplicitDraftTokensBuffers()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers7reshapeE10SizeType3210SizeType32RKN7runtime11ModelConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ExplicitDraftTokensBuffers::reshape()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13setFromInputsE10SizeType3210SizeType32RKN7runtime7ITensorERK7ITensorRKN26ExplicitDraftTokensBuffers6InputsERK7ITensorRKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ExplicitDraftTokensBuffers::setFromInputs()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime26ExplicitDraftTokensBuffers18insertInputTensorsER9TensorMapR9TensorMapRKN7runtime11WorldConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ExplicitDraftTokensBuffers::insertInputTensors()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers12engineInputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ExplicitDraftTokensBuffers::engineInputs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13engineOutputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ExplicitDraftTokensBuffers::engineOutputs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers20scanTempStorageBytesE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ExplicitDraftTokensBuffers::scanTempStorageBytes</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers15scanTempStorageE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ExplicitDraftTokensBuffers::scanTempStorage</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers23cumSumGenerationLengthsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ExplicitDraftTokensBuffers::cumSumGenerationLengths</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0ENK12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13setFromInputsEv10SizeType3210SizeType3210SizeType32RK7ITensorRKN26ExplicitDraftTokensBuffers6InputsERK7ITensorRKN7runtime25ExplicitDraftTokensModuleERKN7runtime10CudaStreamE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ExplicitDraftTokensBuffers::setFromInputs()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers12EngineInputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ExplicitDraftTokensBuffers::EngineInputs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ExplicitDraftTokensBuffers::EngineOutputs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6InputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ExplicitDraftTokensBuffers::Inputs</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#gptdecoder-h">gptDecoder.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm6layersE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::layers</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime20getDefaultBatchSlotsEN7runtime10SizeType32E"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::getDefaultBatchSlots()</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime10GptDecoderE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoder</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder13CudaStreamPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoder::CudaStreamPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder9TensorPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoder::TensorPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder10GptDecoderERKN8executor12DecodingModeE6size_t6size_t6size_t6size_t6size_tRK13CudaStreamPtrNSt10shared_ptrIK25SpeculativeDecodingModuleEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoder::GptDecoder()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder5setupERK14SamplingConfig6size_tRK14TensorConstPtrRKNSt8optionalI14DecodingOutputEERKNSt8optionalIKNSt6vectorIN13decoder_batch7RequestEEEEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoder::setup()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder12forwardAsyncER14DecodingOutputRK13DecodingInput"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoder::forwardAsync()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder11forwardSyncER14DecodingOutputRK13DecodingInput"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoder::forwardSync()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder17getSamplingConfigEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoder::getSamplingConfig()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder16disableLookaheadERKNSt8optionalI14SamplingConfigEE10SizeType3214TensorConstPtr"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoder::disableLookahead()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder8mManagerE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoder::mManager</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder19mDynamicDecodeLayerE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoder::mDynamicDecodeLayer</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder23mDecodingLayerWorkspaceE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoder::mDecodingLayerWorkspace</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder15mSamplingConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoder::mSamplingConfig</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder13mMaxBatchSizeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoder::mMaxBatchSize</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder10mVocabSizeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoder::mVocabSize</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder16mVocabSizePaddedE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoder::mVocabSizePadded</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder13mDecodingModeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoder::mDecodingMode</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11IGptDecoderE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoder</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11IGptDecoder9TensorPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoder::TensorPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11IGptDecoder14TensorConstPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoder::TensorConstPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11IGptDecoderD0Ev"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoder::~IGptDecoder()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11IGptDecoder5setupERK14SamplingConfig6size_tRK14TensorConstPtrRKNSt8optionalI14DecodingOutputEERKNSt8optionalIKNSt6vectorIN13decoder_batch7RequestEEEEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoder::setup()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11IGptDecoder12forwardAsyncER14DecodingOutputRK13DecodingInput"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoder::forwardAsync()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11IGptDecoder11forwardSyncER14DecodingOutputRK13DecodingInput"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoder::forwardSync()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11IGptDecoder17getSamplingConfigEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoder::getSamplingConfig()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11IGptDecoder16disableLookaheadERKNSt8optionalI14SamplingConfigEE10SizeType3214TensorConstPtr"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoder::disableLookahead()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11IGptDecoder6createERKN8executor12DecodingModeEN8nvinfer18DataTypeE6size_t6size_t6size_t6size_t6size_tRKN13BufferManager13CudaStreamPtrERKNSt10shared_ptrIK25SpeculativeDecodingModuleEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoder::create()</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#gptdecoderbatched-h">gptDecoderBatched.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatchedE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched11ForwardTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::ForwardType</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched13CudaStreamPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::CudaStreamPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched13LlmRequestPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::LlmRequestPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched13RequestVectorE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::RequestVector</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9TensorPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::TensorPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14SharedConstPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::SharedConstPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched17GptDecoderBatchedENSt6size_tENSt6size_tE13CudaStreamPtrRK23SpeculativeDecodingModeN8nvinfer18DataTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::GptDecoderBatched()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched5setupERKN8executor12DecodingModeE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType3210SizeType32N8nvinfer18DataTypeERK11ModelConfig"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::setup()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched24setupExplicitDraftTokensEN26ExplicitDraftTokensBuffers6InputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::setupExplicitDraftTokens()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched10setupEagleEN12EagleBuffers6InputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::setupEagle()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14setupLookaheadE24LookaheadDecodingBuffers"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::setupLookahead()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched16disableLookaheadE10SizeType32RK13RequestVector"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::disableLookahead()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched8newBatchERK15GenerationInputRK16GenerationOutputRK14SamplingConfigRK11ModelConfig"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::newBatch()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched11newRequestsERKNSt6vectorI10SizeType32EERKNSt6vectorIN13decoder_batch7RequestEEERKNSt6vectorI14SamplingConfigEERK11ModelConfig"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::newRequests()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched12forwardAsyncERN13decoder_batch6OutputERKN13decoder_batch5InputE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::forwardAsync()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched11forwardSyncERKN13decoder_batch20DecoderFinishedEventE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::forwardSync()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched11forwardSyncERKN13decoder_batch20DecoderFinishedEventERN13decoder_batch6OutputERKN13decoder_batch5InputE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::forwardSync()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched12forwardAsyncERN7decoder6OutputERKN7decoder5InputE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::forwardAsync()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched11forwardSyncEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::forwardSync()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched11getFinishedEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::getFinished()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched16getFinishReasonsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::getFinishReasons()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched6getIdsE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::getIds()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched6getIdsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::getIds()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched14getGatheredIdsE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::getGatheredIds()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched14getGatheredIdsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::getGatheredIds()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched8finalizeE10SizeType32RK14SamplingConfigb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::finalize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched8finalizeERK14SamplingConfig"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::finalize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched12getParentIdsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::getParentIds()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched14getCumLogProbsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::getCumLogProbs()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched14getCumLogProbsE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::getCumLogProbs()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched11getLogProbsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::getLogProbs()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched11getLogProbsE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::getLogProbs()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched15getAllNewTokensEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::getAllNewTokens()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched12getNewTokensE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::getNewTokens()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched10getNbStepsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::getNbSteps()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched13getNbFinishedEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::getNbFinished()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched18getNextDraftTokensEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::getNextDraftTokens()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched25getPrevDraftTokensLengthsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::getPrevDraftTokensLengths()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched25getNextDraftTokensLengthsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::getNextDraftTokensLengths()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched24getAcceptedLengthsCumSumEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::getAcceptedLengthsCumSum()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched22getAcceptedPackedPathsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::getAcceptedPackedPaths()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched15getDecodingModeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::getDecodingMode()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched13GptDecoderPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::GptDecoderPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched16DecodingInputPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::DecodingInputPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched17DecodingOutputPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::DecodingOutputPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched18postProcessRequestE10SizeType32RK14SamplingConfigb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::postProcessRequest()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched10newRequestE10SizeType32RKN13decoder_batch7RequestERK14SamplingConfigRK11ModelConfig"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::newRequest()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched34allocateSpeculativeDecodingBuffersEN8nvinfer18DataTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::allocateSpeculativeDecodingBuffers()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched24setupSpeculativeDecodingERK11ModelConfig"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::setupSpeculativeDecoding()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14setupLookaheadERK11ModelConfig"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::setupLookahead()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched29newRequestSpeculativeDecodingE10SizeType32RKN13decoder_batch7RequestERK14SamplingConfigRK11ModelConfig"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::newRequestSpeculativeDecoding()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched29newRequestDraftTokensExternalE10SizeType32RKN13decoder_batch7RequestERK14SamplingConfig"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::newRequestDraftTokensExternal()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched16newRequestMedusaE10SizeType32RKN13decoder_batch7RequestE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::newRequestMedusa()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched19newRequestLookaheadE10SizeType32RKN13decoder_batch7RequestE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::newRequestLookahead()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched29newRequestExplicitDraftTokensE10SizeType32RKN13decoder_batch7RequestE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::newRequestExplicitDraftTokens()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched15newRequestEagleE10SizeType32RKN13decoder_batch7RequestERK11ModelConfig"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::newRequestEagle()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14updateFinishedERKN13decoder_batch20DecoderFinishedEventE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::updateFinished()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched28setExplicitDraftTokensInputsERKN13decoder_batch5InputE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::setExplicitDraftTokensInputs()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14setEagleInputsERKN13decoder_batch5InputE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::setEagleInputs()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched15forwardDispatchERN13decoder_batch6OutputERKN13decoder_batch5InputE11ForwardType"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::forwardDispatch()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14forwardDecoderE10SizeType32RN13decoder_batch6OutputERKN13decoder_batch5InputE11ForwardType"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::forwardDecoder()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched10mVocabSizeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mVocabSize</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched16mVocabSizePaddedE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mVocabSizePadded</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14mRuntimeStreamE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mRuntimeStream</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14mDecoderStreamE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mDecoderStream</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14mBufferManagerE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mBufferManager</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched19mDecoderFinishEventE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mDecoderFinishEvent</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched13mForwardEventE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mForwardEvent</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched8mDecoderE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mDecoder</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched19mJointDecodingInputE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mJointDecodingInput</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched20mJointDecodingOutputE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mJointDecodingOutput</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched8mNbStepsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mNbSteps</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9mFinishedE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mFinished</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched12mFinishedSumE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mFinishedSum</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched13mMaxNewTokensE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mMaxNewTokens</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched11mBeamWidthsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mBeamWidths</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched24mNumDecodingEngineTokensE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mNumDecodingEngineTokens</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14mFinishedStepsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mFinishedSteps</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched16mBatchSlotsSetupE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mBatchSlotsSetup</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched18mBatchSlotsDecoderE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mBatchSlotsDecoder</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched18mMaxSequenceLengthE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mMaxSequenceLength</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched19mMaxAttentionWindowE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mMaxAttentionWindow</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched16mSinkTokenLengthE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mSinkTokenLength</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched16mActualBatchSizeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mActualBatchSize</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched25mMaxDecodingDecoderTokensE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mMaxDecodingDecoderTokens</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched24mMaxDecodingEngineTokensE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mMaxDecodingEngineTokens</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched24mSpeculativeDecodingModeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mSpeculativeDecodingMode</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched13mDecodingModeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mDecodingMode</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched21mOutputBeamHypothesesE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mOutputBeamHypotheses</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched15mCumLogProbsTmpE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mCumLogProbsTmp</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched7mNumSMsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptDecoderBatched::mNumSMs</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#gptsession-h">gptSession.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm13batch_manager16kv_cache_managerE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::batch_manager::kv_cache_manager</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSessionE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession9LoggerPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::LoggerPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession10GptSessionERK6ConfigRK11ModelConfigRK11WorldConfigRK9RawEngine9LoggerPtr"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::GptSession()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession10GptSessionERK6ConfigRK11ModelConfigRK11WorldConfigPKvNSt6size_tE9LoggerPtr"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::GptSession()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession10GptSessionERK6ConfigRK11ModelConfigRK11WorldConfigRKNSt6vectorI7uint8_tEE9LoggerPtr"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::GptSession()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession10GptSessionERK6ConfigRK11ModelConfigRK11WorldConfigRKNSt6stringE9LoggerPtr"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::GptSession()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession9getLoggerEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::getLogger()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession16getBufferManagerEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::getBufferManager()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession19getRuntimeStreamPtrEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::getRuntimeStreamPtr()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession14getModelConfigEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::getModelConfig()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession14getWorldConfigEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::getWorldConfig()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession9getDeviceEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::getDevice()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession20getNormalizeLogProbsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::getNormalizeLogProbs()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession18getEngineInspectorEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::getEngineInspector()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession16getLogitDataTypeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::getLogitDataType()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession17getTensorDataTypeERKNSt6stringE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::getTensorDataType()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession14getTensorShapeERKNSt6stringE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::getTensorShape()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession8generateER16GenerationOutputRK15GenerationInputRK14SamplingConfigKNSt10shared_ptrI18GenerationProfilerEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::generate()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession16setLayerProfilerEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::setLayerProfiler()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession19getLayerProfileInfoEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::getLayerProfileInfo()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession18BaseKVCacheManagerE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::BaseKVCacheManager</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession13KvCacheConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::KvCacheConfig</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession9TensorPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::TensorPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession22TokenGeneratedCallbackE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::TokenGeneratedCallback</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession13useCudaGraphsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::useCudaGraphs()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession15generateBatchedERNSt6vectorI16GenerationOutputEERKNSt6vectorI15GenerationInputEERK14SamplingConfigRK22TokenGeneratedCallbackKNSt10shared_ptrI18GenerationProfilerEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::generateBatched()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession5setupERK6Config"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::setup()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession14createContextsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::createContexts()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession13createBuffersE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::createBuffers()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession14createDecodersE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType32N8nvinfer18DataTypeEb10SizeType32RKN8executor12DecodingModeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::createDecoders()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession20createKvCacheManagerE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType32RK13KvCacheConfig"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::createKvCacheManager()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession30createCustomAllReduceWorkspaceE10SizeType3210SizeType3210SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::createCustomAllReduceWorkspace()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession18executeContextStepERKNSt6vectorI15GenerationInputEERKNSt6vectorI10SizeType32EEPK18BaseKVCacheManager"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::executeContextStep()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession21executeGenerationStepE10SizeType32RKNSt6vectorI15GenerationInputEERNSt6vectorI16GenerationOutputEERKNSt6vectorI10SizeType32EEP18BaseKVCacheManagerRNSt6vectorIbEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::executeGenerationStep()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession16decoderStepAsyncE10SizeType3210SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::decoderStepAsync()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession14shouldStopSyncE10SizeType3210SizeType3210SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::shouldStopSync()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession8finalizeE10SizeType32RK14SamplingConfig"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::finalize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession19kvCacheAddSequencesE10SizeType3210SizeType3210SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::kvCacheAddSequences()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession11initDecoderER7ITensorRK15GenerationInputRK16GenerationOutputRK14SamplingConfig10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::initDecoder()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession30createOnTokenGeneratedCallbackER16GenerationOutput"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::createOnTokenGeneratedCallback()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession23shouldUseKVCacheManagerEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::shouldUseKVCacheManager()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession12mModelConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::mModelConfig</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession12mWorldConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::mWorldConfig</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession7mDeviceE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::mDevice</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession13mPipelineCommE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::mPipelineComm</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession11mCommStreamE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::mCommStream</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession10mCommEventE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::mCommEvent</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession17mAllReduceBuffersE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::mAllReduceBuffers</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession25mDecoderMaxSequenceLengthE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::mDecoderMaxSequenceLength</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession29mDecoderMaxAttentionWindowVecE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::mDecoderMaxAttentionWindowVec</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession26mDecoderMaxAttentionWindowE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::mDecoderMaxAttentionWindow</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession23mDecoderSinkTokenLengthE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::mDecoderSinkTokenLength</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession7mLoggerE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::mLogger</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession8mRuntimeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::mRuntime</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession15mKvCacheManagerE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::mKvCacheManager</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession17mMicroBatchConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::mMicroBatchConfig</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession9mDecodersE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::mDecoders</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession8mBuffersE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::mBuffers</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession15mReceivedEventsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::mReceivedEvents</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession14mCudaGraphModeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::mCudaGraphMode</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession19mCudaGraphInstancesE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::mCudaGraphInstances</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession18mNormalizeLogProbsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::mNormalizeLogProbs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession6ConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::Config</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession17CudaGraphExecutorE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::CudaGraphExecutor</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession18GenerationProfilerE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::GenerationProfiler</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession16MicroBatchConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::GptSession::MicroBatchConfig</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime5utilsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::utils</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime5utils10loadEngineERKNSt6stringE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::utils::loadEngine()</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#ibuffer-h">iBuffer.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4IEN12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kGPUEEE"><code class="docutils literal notranslate"><span class="pre">MemoryTypeString</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kGPUEE5valueE"><code class="docutils literal notranslate"><span class="pre">MemoryTypeString::value</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4IEN12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kCPUEEE"><code class="docutils literal notranslate"><span class="pre">MemoryTypeString</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kCPUEE5valueE"><code class="docutils literal notranslate"><span class="pre">MemoryTypeString::value</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4IEN12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType7kPINNEDEEE"><code class="docutils literal notranslate"><span class="pre">MemoryTypeString</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType7kPINNEDEE5valueE"><code class="docutils literal notranslate"><span class="pre">MemoryTypeString::value</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4IEN12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kUVMEEE"><code class="docutils literal notranslate"><span class="pre">MemoryTypeString</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kUVMEE5valueE"><code class="docutils literal notranslate"><span class="pre">MemoryTypeString::value</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4IEN12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType11kPINNEDPOOLEEE"><code class="docutils literal notranslate"><span class="pre">MemoryTypeString</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType11kPINNEDPOOLEE5valueE"><code class="docutils literal notranslate"><span class="pre">MemoryTypeString::value</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kFLOATEEE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kFLOATEE4typeE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits::type</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kFLOATEE4nameE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits::name</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kFLOATEE4sizeE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits::size</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kHALFEEE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kHALFEE4typeE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits::type</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kHALFEE4nameE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits::name</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kHALFEE4sizeE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits::size</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kINT8EEE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kINT8EE4typeE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits::type</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kINT8EE4nameE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits::name</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kINT8EE4sizeE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits::size</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EEE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EE4typeE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits::type</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EE4nameE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits::name</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EE4sizeE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits::size</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EEE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EE4typeE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits::type</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EE4nameE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits::name</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EE4sizeE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits::size</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EXL1EEEE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EXL1EEE4typeE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits::type</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EXL1EEE4nameE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits::name</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EXL1EEE4sizeE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits::size</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EXL1EEEE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EXL1EEE4typeE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits::type</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EXL1EEE4nameE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits::name</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EXL1EEE4sizeE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits::size</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4I_bEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kBOOLE9kUnsignedEE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kBOOLE9kUnsignedE4typeE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits::type</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kBOOLE9kUnsignedE4nameE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits::name</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kBOOLE9kUnsignedE4sizeE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits::size</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4I_bEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kUINT8E9kUnsignedEE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kUINT8E9kUnsignedE4typeE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits::type</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kUINT8E9kUnsignedE4nameE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits::name</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kUINT8E9kUnsignedE4sizeE"><code class="docutils literal notranslate"><span class="pre">DataTypeTraits::size</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeINSt6int8_tEEE"><code class="docutils literal notranslate"><span class="pre">TRTDataType</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeINSt6int8_tEE5valueE"><code class="docutils literal notranslate"><span class="pre">TRTDataType::value</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeINSt7int32_tEEE"><code class="docutils literal notranslate"><span class="pre">TRTDataType</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeINSt7int32_tEE5valueE"><code class="docutils literal notranslate"><span class="pre">TRTDataType::value</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeINSt8uint32_tEEE"><code class="docutils literal notranslate"><span class="pre">TRTDataType</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeINSt8uint32_tEE5valueE"><code class="docutils literal notranslate"><span class="pre">TRTDataType::value</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeINSt7int64_tEEE"><code class="docutils literal notranslate"><span class="pre">TRTDataType</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeINSt7int64_tEE5valueE"><code class="docutils literal notranslate"><span class="pre">TRTDataType::value</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeINSt8uint64_tEEE"><code class="docutils literal notranslate"><span class="pre">TRTDataType</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeINSt8uint64_tEE5valueE"><code class="docutils literal notranslate"><span class="pre">TRTDataType::value</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeINSt7uint8_tEEE"><code class="docutils literal notranslate"><span class="pre">TRTDataType</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeINSt7uint8_tEE5valueE"><code class="docutils literal notranslate"><span class="pre">TRTDataType::value</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeIN7kernels12KVCacheIndexEEE"><code class="docutils literal notranslate"><span class="pre">TRTDataType</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeIN7kernels12KVCacheIndexEE5valueE"><code class="docutils literal notranslate"><span class="pre">TRTDataType::value</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeIN7kernels13FinishedStateEEE"><code class="docutils literal notranslate"><span class="pre">TRTDataType</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeIN7kernels13FinishedStateEE5valueE"><code class="docutils literal notranslate"><span class="pre">TRTDataType::value</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeIN7runtime11RequestTypeEEE"><code class="docutils literal notranslate"><span class="pre">TRTDataType</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeIN7runtime11RequestTypeEE5valueE"><code class="docutils literal notranslate"><span class="pre">TRTDataType::value</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime18PointerElementTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::PointerElementType</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryType</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryType4kGPUE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryType::kGPU</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryType4kCPUE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryType::kCPU</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryType7kPINNEDE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryType::kPINNED</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryType4kUVME"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryType::kUVM</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryType11kPINNEDPOOLE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryType::kPINNEDPOOL</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime16constPointerCastENSt10shared_ptrINSt14remove_const_tI1TEEEERKNSt10shared_ptrI1TEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::constPointerCast()</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4I00EN12tensorrt_llm7runtime16constPointerCastENSt10shared_ptrINSt14remove_const_tI1TEEEERRNSt10unique_ptrI1T1DEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::constPointerCast()</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime10bufferCastEPK1TRK7IBuffer"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::bufferCast()</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime10bufferCastEP1TR7IBuffer"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::bufferCast()</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEP1TRKN7IBuffer9SharedPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::bufferCastOrNull()</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEPK1TRKN7IBuffer14SharedConstPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::bufferCastOrNull()</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEP1TRKNSt8optionalIN7IBuffer9SharedPtrEEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::bufferCastOrNull()</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEPK1TRKNSt8optionalIN7IBuffer14SharedConstPtrEEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::bufferCastOrNull()</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtimelsERNSt7ostreamERK7IBuffer"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::operator<<()</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14BufferDataTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::BufferDataType</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14BufferDataType14BufferDataTypeEN8nvinfer18DataTypeEbb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::BufferDataType::BufferDataType()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime14BufferDataTypecvN8nvinfer18DataTypeEEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::BufferDataType::operator</span> <span class="pre">nvinfer1::DataType()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime14BufferDataType11getDataTypeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::BufferDataType::getDataType()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime14BufferDataType9isPointerEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::BufferDataType::isPointer()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime14BufferDataType10isUnsignedEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::BufferDataType::isUnsigned()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime14BufferDataType7getSizeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::BufferDataType::getSize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime14BufferDataType13getSizeInBitsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::BufferDataType::getSizeInBits()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14BufferDataType15kTrtPointerTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::BufferDataType::kTrtPointerType</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14BufferDataType9mDataTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::BufferDataType::mDataType</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14BufferDataType9mUnsignedE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::BufferDataType::mUnsigned</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14BufferDataType8mPointerE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::BufferDataType::mPointer</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime11BufferRangeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::BufferRange</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange4BaseE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::BufferRange::Base</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange11BufferRangeEP1T9size_type"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::BufferRange::BufferRange()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tIXntNSt10is_const_vI1UEEEbEEEN12tensorrt_llm7runtime11BufferRange11BufferRangeER7IBuffer"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::BufferRange::BufferRange()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI1UEEbEEEN12tensorrt_llm7runtime11BufferRange11BufferRangeERK7IBuffer"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::BufferRange::BufferRange()</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4I_N8nvinfer18DataTypeE_b_bEN12tensorrt_llm7runtime14DataTypeTraitsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DataTypeTraits</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4I_N8nvinfer18DataTypeE_bEN12tensorrt_llm7runtime14DataTypeTraitsI9kDataType9kUnsignedXL1EEEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DataTypeTraits</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsI9kDataType9kUnsignedXL1EEE4typeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DataTypeTraits::type</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsI9kDataType9kUnsignedXL1EEE4nameE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DataTypeTraits::name</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsI9kDataType9kUnsignedXL1EEE4sizeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::DataTypeTraits::size</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19PhonyNameDueToError4typeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::PhonyNameDueToError::type</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19PhonyNameDueToError4nameE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::PhonyNameDueToError::name</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19PhonyNameDueToError4sizeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::PhonyNameDueToError::size</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer9UniquePtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::UniquePtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer9SharedPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::SharedPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer14UniqueConstPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::UniqueConstPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer14SharedConstPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::SharedConstPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer8DataTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::DataType</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer4dataEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::data()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime7IBuffer4dataEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::data()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer4dataENSt6size_tE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::data()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime7IBuffer4dataENSt6size_tE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::data()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime7IBuffer7getSizeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::getSize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime7IBuffer14getSizeInBytesEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::getSizeInBytes()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime7IBuffer11getCapacityEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::getCapacity()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime7IBuffer11getDataTypeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::getDataType()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime7IBuffer15getDataTypeNameEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::getDataTypeName()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime7IBuffer13getMemoryTypeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::getMemoryType()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime7IBuffer17getMemoryTypeNameEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::getMemoryTypeName()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer6resizeENSt6size_tE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::resize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer7releaseEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::release()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferD0Ev"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::~IBuffer()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer7IBufferERK7IBuffer"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::IBuffer()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferaSERK7IBuffer"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::operator=()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer15getDataTypeNameE8DataType"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::getDataTypeName()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer5sliceE9SharedPtrNSt6size_tENSt6size_tE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::slice()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7IBuffer5sliceE14UniqueConstPtrRR9TConstPtrNSt6size_tENSt6size_tE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::slice()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer5sliceE9SharedPtrNSt6size_tE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::slice()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7IBuffer5sliceE14UniqueConstPtrRR9TConstPtrNSt6size_tE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::slice()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer4viewE9SharedPtr"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::view()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer4viewE9SharedPtrNSt6size_tE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::view()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7IBuffer4viewE14UniqueConstPtrRR9TConstPtrNSt6size_tE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::view()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer4wrapEPv8DataTypeNSt6size_tENSt6size_tE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::wrap()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer4wrapEPv8DataTypeNSt6size_tE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::wrap()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime7IBuffer4wrapE9UniquePtrP1TNSt6size_tENSt6size_tE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::wrap()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime7IBuffer4wrapE9UniquePtrP1TNSt6size_tE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::wrap()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime7IBuffer4wrapE9UniquePtrRNSt6vectorI1TEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::wrap()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer10memoryTypeEPKv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::memoryType()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer7IBufferEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::IBuffer()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime7IBuffer7toBytesENSt6size_tE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IBuffer::toBytes()</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4I_10MemoryTypeEN12tensorrt_llm7runtime16MemoryTypeStringE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::MemoryTypeString</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19PhonyNameDueToError5valueE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::PhonyNameDueToError::value</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4I0_bEN12tensorrt_llm7runtime11TRTDataTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::TRTDataType</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeIbEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::TRTDataType</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeIbE5valueE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::TRTDataType::value</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeIfEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::TRTDataType</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeIfE5valueE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::TRTDataType::value</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeI4halfEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::TRTDataType</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeI4halfE5valueE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::TRTDataType::value</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime11TRTDataTypeIP1TEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::TRTDataType</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeIP1TE5valueE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::TRTDataType::value</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeIP1TE15kUnderlyingTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::TRTDataType::kUnderlyingType</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeIPvEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::TRTDataType</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeIPvE5valueE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::TRTDataType::value</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#igptdecoderbatched-h">iGptDecoderBatched.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatchedE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched13CudaStreamPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::CudaStreamPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched13LlmRequestPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::LlmRequestPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched13RequestVectorE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::RequestVector</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched9TensorPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::TensorPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched23DecoderFinishedEventPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::DecoderFinishedEventPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched24setupExplicitDraftTokensEN26ExplicitDraftTokensBuffers6InputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::setupExplicitDraftTokens()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched10setupEagleEN12EagleBuffers6InputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::setupEagle()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched14setupLookaheadE24LookaheadDecodingBuffers"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::setupLookahead()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched16disableLookaheadE10SizeType32RK13RequestVector"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::disableLookahead()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched12forwardAsyncERN13decoder_batch6OutputERKN13decoder_batch5InputE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::forwardAsync()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched11forwardSyncERKN13decoder_batch20DecoderFinishedEventERN13decoder_batch6OutputERKN13decoder_batch5InputE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::forwardSync()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched11forwardSyncERKN13decoder_batch20DecoderFinishedEventE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::forwardSync()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched7forwardERN13decoder_batch6OutputERKN13decoder_batch5InputE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::forward()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched6getIdsE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::getIds()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched14getGatheredIdsE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::getGatheredIds()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched8finalizeE10SizeType32RK14SamplingConfigb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::finalize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched11getFinishedEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::getFinished()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched16getFinishReasonsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::getFinishReasons()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched14getCumLogProbsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::getCumLogProbs()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched14getCumLogProbsE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::getCumLogProbs()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched11getLogProbsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::getLogProbs()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched11getLogProbsE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::getLogProbs()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched12getParentIdsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::getParentIds()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched10getNbStepsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::getNbSteps()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched15getDecodingModeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::getDecodingMode()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched11newRequestsERKNSt6vectorI10SizeType32EERKNSt6vectorIN13decoder_batch7RequestEEERKNSt6vectorI14SamplingConfigEERK11ModelConfig"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::newRequests()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched18getNextDraftTokensEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::getNextDraftTokens()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched25getPrevDraftTokensLengthsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::getPrevDraftTokensLengths()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched25getNextDraftTokensLengthsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::getNextDraftTokensLengths()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched24getAcceptedLengthsCumSumEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::getAcceptedLengthsCumSum()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched22getAcceptedPackedPathsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::getAcceptedPackedPaths()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched18IGptDecoderBatchedEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IGptDecoderBatched::IGptDecoderBatched()</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch6OutputE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::decoder_batch::Output</span></code></a></li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch20DecoderFinishedEventE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::decoder_batch::DecoderFinishedEvent</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch20DecoderFinishedEvent20DecoderFinishedEventERR9CudaEventRKNSt6vectorIbEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::decoder_batch::DecoderFinishedEvent::DecoderFinishedEvent()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch20DecoderFinishedEvent5eventE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::decoder_batch::DecoderFinishedEvent::event</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch20DecoderFinishedEvent6activeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::decoder_batch::DecoderFinishedEvent::active</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5InputE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::decoder_batch::Input</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input14TensorConstPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::decoder_batch::Input::TensorConstPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input9TensorPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::decoder_batch::Input::TensorPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input5InputERKNSt6vectorI9TensorPtrEERKNSt6vectorIbEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::decoder_batch::Input::Input()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input5InputERKNSt6vectorI9TensorPtrEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::decoder_batch::Input::Input()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input6logitsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::decoder_batch::Input::logits</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input6activeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::decoder_batch::Input::active</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input16cacheIndirectionE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::decoder_batch::Input::cacheIndirection</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input20predictedDraftLogitsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::decoder_batch::Input::predictedDraftLogits</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input8seqSlotsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::decoder_batch::Input::seqSlots</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input25explicitDraftTokensInputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::decoder_batch::Input::explicitDraftTokensInputs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input29explicitDraftTokensLastInputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::decoder_batch::Input::explicitDraftTokensLastInputs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input11eagleInputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::decoder_batch::Input::eagleInputs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input15eagleLastInputsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::decoder_batch::Input::eagleLastInputs</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#istatefulgptdecoder-h">iStatefulGptDecoder.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoderE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IStatefulGptDecoder</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder13CudaStreamPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IStatefulGptDecoder::CudaStreamPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder9TensorPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IStatefulGptDecoder::TensorPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder5setupERKN8executor12DecodingModeE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType3210SizeType32N8nvinfer18DataTypeERK11ModelConfig"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IStatefulGptDecoder::setup()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder8newBatchERK15GenerationInputRK16GenerationOutputRK14SamplingConfigRK11ModelConfig"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IStatefulGptDecoder::newBatch()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder12forwardAsyncERN7decoder6OutputERKN7decoder5InputE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IStatefulGptDecoder::forwardAsync()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder11forwardSyncEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IStatefulGptDecoder::forwardSync()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder7forwardERN7decoder6OutputERKN7decoder5InputE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IStatefulGptDecoder::forward()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime19IStatefulGptDecoder8finalizeERK14SamplingConfig"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IStatefulGptDecoder::finalize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime19IStatefulGptDecoder6getIdsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IStatefulGptDecoder::getIds()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime19IStatefulGptDecoder14getGatheredIdsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IStatefulGptDecoder::getGatheredIds()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime19IStatefulGptDecoder14getCumLogProbsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IStatefulGptDecoder::getCumLogProbs()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime19IStatefulGptDecoder11getLogProbsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IStatefulGptDecoder::getLogProbs()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime19IStatefulGptDecoder12getNewTokensE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IStatefulGptDecoder::getNewTokens()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime19IStatefulGptDecoder15getAllNewTokensEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IStatefulGptDecoder::getAllNewTokens()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime19IStatefulGptDecoder13getNbFinishedEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IStatefulGptDecoder::getNbFinished()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoderD0Ev"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IStatefulGptDecoder::~IStatefulGptDecoder()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder19IStatefulGptDecoderEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::IStatefulGptDecoder::IStatefulGptDecoder()</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7decoderE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::decoder</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7decoder5InputE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::decoder::Input</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7decoder6OutputE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::decoder::Output</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#lookaheadbuffers-h">lookaheadBuffers.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffersE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadDecodingBuffers</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers10SizeType32E"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadDecodingBuffers::SizeType32</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers9TensorPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadDecodingBuffers::TensorPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers7ITensorE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadDecodingBuffers::ITensor</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers24LookaheadDecodingBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadDecodingBuffers::LookaheadDecodingBuffers()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers17generationLengthsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadDecodingBuffers::generationLengths</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers15positionOffsetsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadDecodingBuffers::positionOffsets</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers11packedMasksE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadDecodingBuffers::packedMasks</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers11positionIdsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadDecodingBuffers::positionIds</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffersE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadRuntimeBuffers</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers10SizeType32E"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadRuntimeBuffers::SizeType32</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers7ITensorE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadRuntimeBuffers::ITensor</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers9TensorPtrE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadRuntimeBuffers::TensorPtr</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers9TensorMapE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadRuntimeBuffers::TensorMap</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers23LookaheadRuntimeBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerERKN7runtime11ModelConfigERKN7runtime11WorldConfigERKN8executor14DecodingConfigERKN7runtime11TllmRuntimeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadRuntimeBuffers::LookaheadRuntimeBuffers()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime23LookaheadRuntimeBuffers13setFromInputsE10SizeType3210SizeType32RKN7runtime7ITensorERK7ITensorRK24LookaheadDecodingBuffersRKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadRuntimeBuffers::setFromInputs()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers7reshapeE10SizeType3210SizeType3210SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadRuntimeBuffers::reshape()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime23LookaheadRuntimeBuffers18insertInputTensorsER9TensorMapR9TensorMapRKN7runtime11WorldConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadRuntimeBuffers::insertInputTensors()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers23enableLookaheadDecodingE10SizeType3210SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadRuntimeBuffers::enableLookaheadDecoding()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers24disableLookaheadDecodingEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadRuntimeBuffers::disableLookaheadDecoding()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers12cumSumLengthE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadRuntimeBuffers::cumSumLength</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers17packedMasksDeviceE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadRuntimeBuffers::packedMasksDevice</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers23generationLengthsDeviceE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadRuntimeBuffers::generationLengthsDevice</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers21positionOffsetsDeviceE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadRuntimeBuffers::positionOffsetsDevice</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers17positionIdsDeviceE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadRuntimeBuffers::positionIdsDevice</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers14packedMaskHostE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadRuntimeBuffers::packedMaskHost</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers21generationLengthsHostE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadRuntimeBuffers::generationLengthsHost</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers19positionOffsetsHostE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadRuntimeBuffers::positionOffsetsHost</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers15positionIdsHostE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadRuntimeBuffers::positionIdsHost</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers18packedMaskHostCopyE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadRuntimeBuffers::packedMaskHostCopy</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers25generationLengthsHostCopyE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadRuntimeBuffers::generationLengthsHostCopy</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers23positionOffsetsHostCopyE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadRuntimeBuffers::positionOffsetsHostCopy</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers19positionIdsHostCopyE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadRuntimeBuffers::positionIdsHostCopy</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers15useSpecDecodingE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadRuntimeBuffers::useSpecDecoding</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers18batchSlotsHostCopyE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadRuntimeBuffers::batchSlotsHostCopy</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#lookaheadmodule-h">lookaheadModule.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15LookaheadModuleE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadModule</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15LookaheadModule15LookaheadModuleE10SizeType3210SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadModule::LookaheadModule()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15LookaheadModule15LookaheadModuleEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadModule::LookaheadModule()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15LookaheadModule18setExecutionConfigERKN8executor23LookaheadDecodingConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadModule::setExecutionConfig()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime15LookaheadModule18getExecutionConfigEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadModule::getExecutionConfig()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15LookaheadModule16mExecutionConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::LookaheadModule::mExecutionConfig</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#modelconfig-h">modelConfig.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12ModelVariantE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::ModelVariant</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig9LayerTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::LayerType</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig11KVCacheTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::KVCacheType</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig17ManageWeightsTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::ManageWeightsType</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig11ModelConfigE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType3210SizeType32N8nvinfer18DataTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::ModelConfig()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig12getVocabSizeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getVocabSize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig18getVocabSizePaddedE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getVocabSizePadded()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig16countLocalLayersE9LayerType10SizeType3210SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::countLocalLayers()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig20countLowerRankLayersE9LayerType10SizeType3210SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::countLowerRankLayers()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig11getNbLayersE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getNbLayers()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig20getNbAttentionLayersE10SizeType3210SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getNbAttentionLayers()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig14getNbRnnLayersE10SizeType3210SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getNbRnnLayers()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig10getNbHeadsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getNbHeads()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig12getNbKvHeadsE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getNbKvHeads()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12setNbKvHeadsE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setNbKvHeads()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig17setNbCrossKvHeadsE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setNbCrossKvHeads()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig13getHiddenSizeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getHiddenSize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig20getEncoderHiddenSizeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getEncoderHiddenSize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig20setEncoderHiddenSizeE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setEncoderHiddenSize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig14getSizePerHeadEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getSizePerHead()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig14setSizePerHeadE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setSizePerHead()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig11getDataTypeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getDataType()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig21useGptAttentionPluginEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::useGptAttentionPlugin()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig21useGptAttentionPluginEb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::useGptAttentionPlugin()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig20useMambaConv1dPluginEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::useMambaConv1dPlugin()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig20useMambaConv1dPluginEb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::useMambaConv1dPlugin()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig14usePackedInputEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::usePackedInput()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig14usePackedInputEb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::usePackedInput()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig13usePagedStateEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::usePagedState()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig13usePagedStateEb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::usePagedState()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig17getTokensPerBlockEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getTokensPerBlock()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig17setTokensPerBlockE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setTokensPerBlock()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig12getQuantModeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getQuantMode()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12setQuantModeEN6common9QuantModeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setQuantMode()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig24supportsInflightBatchingEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::supportsInflightBatching()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig15getMaxBatchSizeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getMaxBatchSize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig15setMaxBatchSizeE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setMaxBatchSize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig15getMaxBeamWidthEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getMaxBeamWidth()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig15setMaxBeamWidthE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setMaxBeamWidth()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig14getMaxInputLenEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getMaxInputLen()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig14setMaxInputLenE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setMaxInputLen()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig17getMaxSequenceLenEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getMaxSequenceLen()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig17setMaxSequenceLenE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setMaxSequenceLen()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig15getMaxNumTokensEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getMaxNumTokens()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig15setMaxNumTokensENSt8optionalI10SizeType32EE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setMaxNumTokens()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig16getMaxEncoderLenEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getMaxEncoderLen()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig16setMaxEncoderLenE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setMaxEncoderLen()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig15usePromptTuningEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::usePromptTuning()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig8useMropeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::useMrope()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig11setUseMropeEb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setUseMrope()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig24getMaxPositionEmbeddingsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getMaxPositionEmbeddings()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig24setMaxPositionEmbeddingsE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setMaxPositionEmbeddings()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig21getRotaryEmbeddingDimEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getRotaryEmbeddingDim()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig21setRotaryEmbeddingDimE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setRotaryEmbeddingDim()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig30getMaxPromptEmbeddingTableSizeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getMaxPromptEmbeddingTableSize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig30setMaxPromptEmbeddingTableSizeE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setMaxPromptEmbeddingTableSize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig20computeContextLogitsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::computeContextLogits()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig20computeContextLogitsEb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::computeContextLogits()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig23computeGenerationLogitsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::computeGenerationLogits()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig23computeGenerationLogitsEb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::computeGenerationLogits()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig15getModelVariantEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getModelVariant()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig15setModelVariantE12ModelVariant"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setModelVariant()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig25getMaxDecodingDraftTokensEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getMaxDecodingDraftTokens()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig20getMaxDecodingTokensEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getMaxDecodingTokens()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig14setContextFMHAEb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setContextFMHA()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig14getContextFMHAEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getContextFMHA()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig19setPagedContextFMHAEb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setPagedContextFMHA()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig19getPagedContextFMHAEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getPagedContextFMHA()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig18setPpReduceScatterEb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setPpReduceScatter()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig18getPpReduceScatterEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getPpReduceScatter()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig13useLoraPluginEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::useLoraPlugin()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig13useLoraPluginEb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::useLoraPlugin()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig14getLoraModulesEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getLoraModules()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig14setLoraModulesERKNSt6vectorI10LoraModuleEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setLoraModules()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig16getMlpHiddenSizeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getMlpHiddenSize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig16setMlpHiddenSizeE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setMlpHiddenSize()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig16isKVCacheEnabledEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::isKVCacheEnabled()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig14isPagedKVCacheEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::isPagedKVCache()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig19isContinuousKVCacheEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::isContinuousKVCache()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig14getKVCacheTypeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getKVCacheType()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig14setKVCacheTypeE11KVCacheType"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setKVCacheType()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig17useCrossAttentionEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::useCrossAttention()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig20setUseCrossAttentionEb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setUseCrossAttention()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig20usePositionEmbeddingEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::usePositionEmbedding()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig23setUsePositionEmbeddingEb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setUsePositionEmbedding()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig21useTokenTypeEmbeddingEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::useTokenTypeEmbedding()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig24setUseTokenTypeEmbeddingEb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setUseTokenTypeEmbedding()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig14getMaxLoraRankEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getMaxLoraRank()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig14setMaxLoraRankE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setMaxLoraRank()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig26setSpeculativeDecodingModeE23SpeculativeDecodingMode"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setSpeculativeDecodingMode()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig28hasSpeculativeDecodingModuleEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::hasSpeculativeDecodingModule()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig28getSpeculativeDecodingModuleEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getSpeculativeDecodingModule()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig31getSpeculativeDecodingModulePtrEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getSpeculativeDecodingModulePtr()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig31getSpeculativeDecodingModulePtrEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getSpeculativeDecodingModulePtr()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig28setSpeculativeDecodingModuleERKNSt10shared_ptrI25SpeculativeDecodingModuleEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setSpeculativeDecodingModule()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig30resetSpeculativeDecodingModuleEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::resetSpeculativeDecodingModule()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig31enableSeamlessLookaheadDecodingE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::enableSeamlessLookaheadDecoding()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig32disableSeamlessLookaheadDecodingEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::disableSeamlessLookaheadDecoding()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig13getKvDataTypeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getKvDataType()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig18isTransformerBasedEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::isTransformerBased()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig12hasRnnConfigEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::hasRnnConfig()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig12getRnnConfigEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getRnnConfig()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12setRnnConfigERK9RnnConfig"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setRnnConfig()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig10isRnnBasedEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::isRnnBased()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig13getLayerTypesEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getLayerTypes()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig13setLayerTypesERKNSt6vectorI9LayerTypeEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setLayerTypes()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig26getSpeculativeDecodingModeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getSpeculativeDecodingMode()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig14setLogitsDtypeEN8nvinfer18DataTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setLogitsDtype()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig14getLogitsDtypeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getLogitsDtype()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig20setUseShapeInferenceEb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setUseShapeInference()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig17useShapeInferenceEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::useShapeInference()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig20getManageWeightsTypeEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getManageWeightsType()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig20setManageWeightsTypeEK17ManageWeightsType"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setManageWeightsType()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig12getModelNameEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getModelName()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12setModelNameERKNSt6stringE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setModelName()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig21getNumKvHeadsPerLayerEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getNumKvHeadsPerLayer()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig31getNumKvHeadsPerLayerLocalRangeE10SizeType3210SizeType32b"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getNumKvHeadsPerLayerLocalRange()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig21setNumKvHeadsPerLayerERKNSt6vectorI10SizeType32EE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setNumKvHeadsPerLayer()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig26setNumKvHeadsPerCrossLayerERKNSt6vectorI10SizeType32EE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setNumKvHeadsPerCrossLayer()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig18getSumLocalKvHeadsE10SizeType3210SizeType32b"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getSumLocalKvHeads()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig19skipCrossAttnBlocksEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::skipCrossAttnBlocks()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig22setSkipCrossAttnBlocksEb"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::setSkipCrossAttnBlocks()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig21KVCacheTypeFromStringENSt6stringE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::KVCacheTypeFromString()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig25getOptProfilesSplitPointsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::getOptProfilesSplitPoints()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig26kOPT_PROFILES_SPLIT_POINTSE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::kOPT_PROFILES_SPLIT_POINTS</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig29kDEFAULT_NUM_TOKENS_PER_BLOCKE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::kDEFAULT_NUM_TOKENS_PER_BLOCK</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig10mVocabSizeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mVocabSize</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig9mNbLayersE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mNbLayers</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig18mNbAttentionLayersE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mNbAttentionLayers</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12mNbRnnLayersE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mNbRnnLayers</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig8mNbHeadsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mNbHeads</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig11mHiddenSizeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mHiddenSize</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12mSizePerHeadE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mSizePerHead</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig9mDataTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mDataType</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig22mUseGptAttentionPluginE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mUseGptAttentionPlugin</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig21mUseMambaConv1dPluginE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mUseMambaConv1dPlugin</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12mInputPackedE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mInputPacked</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig11mPagedStateE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mPagedState</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig15mTokensPerBlockE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mTokensPerBlock</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig10mQuantModeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mQuantMode</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig13mMaxBatchSizeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mMaxBatchSize</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig13mMaxBeamWidthE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mMaxBeamWidth</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12mMaxInputLenE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mMaxInputLen</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig15mMaxSequenceLenE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mMaxSequenceLen</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig13mMaxNumTokensE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mMaxNumTokens</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig21mComputeContextLogitsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mComputeContextLogits</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig24mComputeGenerationLogitsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mComputeGenerationLogits</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig13mModelVariantE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mModelVariant</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig28mMaxPromptEmbeddingTableSizeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mMaxPromptEmbeddingTableSize</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig9mUseMropeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mUseMrope</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig22mMaxPositionEmbeddingsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mMaxPositionEmbeddings</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig19mRotaryEmbeddingDimE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mRotaryEmbeddingDim</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12mContextFMHAE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mContextFMHA</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig17mPagedContextFMHAE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mPagedContextFMHA</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig16mPpReduceScatterE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mPpReduceScatter</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig14mUseLoraPluginE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mUseLoraPlugin</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12mLoraModulesE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mLoraModules</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig14mMlpHiddenSizeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mMlpHiddenSize</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12mMaxLoraRankE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mMaxLoraRank</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig10mRnnConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mRnnConfig</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12mKVCacheTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mKVCacheType</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig14mMaxEncoderLenE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mMaxEncoderLen</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig18mEncoderHiddenSizeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mEncoderHiddenSize</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig18mUseCrossAttentionE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mUseCrossAttention</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig21mUsePositionEmbeddingE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mUsePositionEmbedding</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig22mUseTokenTypeEmbeddingE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mUseTokenTypeEmbedding</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig11mLayerTypesE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mLayerTypes</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig26mSpeculativeDecodingModuleE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mSpeculativeDecodingModule</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig24mSpeculativeDecodingModeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mSpeculativeDecodingMode</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12mLogitsDtypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mLogitsDtype</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig18mUseShapeInferenceE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mUseShapeInference</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig18mManageWeightsTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mManageWeightsType</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig10mModelNameE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mModelName</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig28mNumKvHeadsPerAttentionLayerE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mNumKvHeadsPerAttentionLayer</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig33mNumKvHeadsPerCrossAttentionLayerE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mNumKvHeadsPerCrossAttentionLayer</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig20mSkipCrossAttnBlocksE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::mSkipCrossAttnBlocks</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig9RnnConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::ModelConfig::RnnConfig</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#samplingconfig-h">samplingConfig.h</a><ul>
|
||
<li class="toctree-l3"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfigE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig</span></code></a><ul>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig14SamplingConfigE10SizeType32"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::SamplingConfig()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig14SamplingConfigERKNSt6vectorI14SamplingConfigEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::SamplingConfig()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig14SamplingConfigERKN8executor14SamplingConfigERKNSt8optionalIN8executor25ExternalDraftTokensConfigEEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::SamplingConfig()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig8validateEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::validate()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig16useDefaultValuesEbRK6OptVecI1TE1T"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::useDefaultValues()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime14SamplingConfigeqERK14SamplingConfig"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::operator==()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime14SamplingConfig17getNumReturnBeamsEv"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::getNumReturnBeams()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig9beamWidthE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::beamWidth</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig18numReturnSequencesE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::numReturnSequences</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig11temperatureE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::temperature</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig19originalTemperatureE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::originalTemperature</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig9minLengthE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::minLength</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig17repetitionPenaltyE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::repetitionPenalty</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig15presencePenaltyE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::presencePenalty</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig16frequencyPenaltyE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::frequencyPenalty</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig17noRepeatNgramSizeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::noRepeatNgramSize</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig14outputLogProbsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::outputLogProbs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig11cumLogProbsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::cumLogProbs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig4topKE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::topK</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig4topPE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::topP</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig10randomSeedE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::randomSeed</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig9topPDecayE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::topPDecay</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig7topPMinE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::topPMin</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig12topPResetIdsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::topPResetIds</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig23beamSearchDiversityRateE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::beamSearchDiversityRate</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig13lengthPenaltyE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::lengthPenalty</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig13earlyStoppingE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::earlyStopping</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig24draftAcceptanceThresholdE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::draftAcceptanceThreshold</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig15topKMedusaHeadsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::topKMedusaHeads</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig17normalizeLogProbsE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::normalizeLogProbs</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig9FloatTypeE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::FloatType</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::OptVec</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig3VecE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::Vec</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig11validateVecEbNSt6stringERK6OptVecI1TE1TNSt8optionalI1TEE"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::validateVec()</span></code></a></li>
|
||
<li class="toctree-l4"><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig10fuseValuesE6OptVecI1TERKNSt6vectorI14SamplingConfigEENSt8functionIF6OptVecI1TE6size_tEEE1T"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm::runtime::SamplingConfig::fuseValues()</span></code></a></li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
</ul>
|
||
<p class="caption" role="heading"><span class="caption-text">Command-Line Reference</span></p>
|
||
<ul>
|
||
<li class="toctree-l1"><a class="reference internal" href="../commands/trtllm-build.html">trtllm-build</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../commands/trtllm-serve.html">trtllm-serve</a></li>
|
||
</ul>
|
||
<p class="caption" role="heading"><span class="caption-text">Architecture</span></p>
|
||
<ul>
|
||
<li class="toctree-l1"><a class="reference internal" href="../architecture/overview.html">TensorRT-LLM Architecture</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../architecture/core-concepts.html">Model Definition</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../architecture/core-concepts.html#compilation">Compilation</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../architecture/core-concepts.html#runtime">Runtime</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../architecture/core-concepts.html#multi-gpu-and-multi-node-support">Multi-GPU and Multi-Node Support</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../architecture/checkpoint.html">TensorRT-LLM Checkpoint</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../architecture/workflow.html">TensorRT-LLM Build Workflow</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../architecture/add-model.html">Adding a Model</a></li>
|
||
</ul>
|
||
<p class="caption" role="heading"><span class="caption-text">Advanced</span></p>
|
||
<ul>
|
||
<li class="toctree-l1"><a class="reference internal" href="../advanced/gpt-attention.html">Multi-Head, Multi-Query, and Group-Query Attention</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../advanced/gpt-runtime.html">C++ GPT Runtime</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../advanced/executor.html">Executor API</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../advanced/graph-rewriting.html">Graph Rewriting Module</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../advanced/inference-request.html">Inference Request</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../advanced/inference-request.html#responses">Responses</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../advanced/lora.html">Run gpt-2b + LoRA using GptManager / cpp runtime</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../advanced/expert-parallelism.html">Expert Parallelism in TensorRT-LLM</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../advanced/kv-cache-reuse.html">KV cache reuse</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../advanced/speculative-decoding.html">Speculative Sampling</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../advanced/disaggregated-service.html">Disaggregated-Service (experimental)</a></li>
|
||
</ul>
|
||
<p class="caption" role="heading"><span class="caption-text">Performance</span></p>
|
||
<ul>
|
||
<li class="toctree-l1"><a class="reference internal" href="../performance/perf-overview.html">Overview</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../performance/perf-benchmarking.html">Benchmarking</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../performance/performance-tuning-guide/index.html">Performance Tuning Guide</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../performance/perf-analysis.html">Performance Analysis</a></li>
|
||
</ul>
|
||
<p class="caption" role="heading"><span class="caption-text">Reference</span></p>
|
||
<ul>
|
||
<li class="toctree-l1"><a class="reference internal" href="../reference/troubleshooting.html">Troubleshooting</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../reference/support-matrix.html">Support Matrix</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../reference/precision.html">Numerical Precision</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../reference/memory.html">Memory Usage of TensorRT-LLM</a></li>
|
||
</ul>
|
||
<p class="caption" role="heading"><span class="caption-text">Blogs</span></p>
|
||
<ul>
|
||
<li class="toctree-l1"><a class="reference internal" href="../blogs/H100vsA100.html">H100 has 4.6x A100 Performance in TensorRT-LLM, achieving 10,000 tok/s at 100ms to first token</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../blogs/H200launch.html">H200 achieves nearly 12,000 tokens/sec on Llama2-13B with TensorRT-LLM</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../blogs/Falcon180B-H200.html">Falcon-180B on a single H200 GPU with INT4 AWQ, and 6.7x faster Llama-70B over A100</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../blogs/quantization-in-TRT-LLM.html">Speed up inference with SOTA quantization techniques in TRT-LLM</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../blogs/XQA-kernel.html">New XQA-kernel provides 2.4x more Llama-70B throughput within the same latency budget</a></li>
|
||
</ul>
|
||
|
||
</div>
|
||
</div>
|
||
</nav>
|
||
|
||
<section data-toggle="wy-nav-shift" class="wy-nav-content-wrap"><nav class="wy-nav-top" aria-label="Mobile navigation menu" >
|
||
<i data-toggle="wy-nav-top" class="fa fa-bars"></i>
|
||
<a href="../index.html">tensorrt_llm</a>
|
||
</nav>
|
||
|
||
<div class="wy-nav-content">
|
||
<div class="rst-content">
|
||
<div role="navigation" aria-label="Page navigation">
|
||
<ul class="wy-breadcrumbs">
|
||
<li><a href="../index.html" class="icon icon-home" aria-label="Home"></a></li>
|
||
<li class="breadcrumb-item active">Runtime</li>
|
||
<li class="wy-breadcrumbs-aside">
|
||
<a href="../_sources/_cpp_gen/runtime.rst.txt" rel="nofollow"> View page source</a>
|
||
</li>
|
||
</ul>
|
||
<hr/>
|
||
</div>
|
||
<div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
|
||
<div itemprop="articleBody">
|
||
|
||
<section id="runtime">
|
||
<h1>Runtime<a class="headerlink" href="#runtime" title="Link to this heading"></a></h1>
|
||
<section id="buffermanager-h">
|
||
<h2>bufferManager.h<a class="headerlink" href="#buffermanager-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv412tensorrt_llm">
|
||
<span id="_CPPv312tensorrt_llm"></span><span id="_CPPv212tensorrt_llm"></span><span id="tensorrt_llm"></span><span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtimeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtimeE"></span><span id="_CPPv2N12tensorrt_llm7runtimeE"></span><span id="tensorrt_llm::runtime"></span><span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13BufferManagerE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13BufferManagerE"></span><span id="_CPPv2N12tensorrt_llm7runtime13BufferManagerE"></span><span id="tensorrt_llm::runtime::BufferManager"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">BufferManager</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13BufferManagerE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="docutils container">
|
||
<em>#include <bufferManager.h></em></div>
|
||
<p>A helper class for managing memory on host and device. </p>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13BufferManager10IBufferPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13BufferManager10IBufferPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime13BufferManager10IBufferPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a0bd5b422b1e3a5d2171f66532f27f413"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">IBufferPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer9UniquePtrE" title="tensorrt_llm::runtime::IBuffer::UniquePtr"><span class="n"><span class="pre">UniquePtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10IBufferPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13BufferManager10ITensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13BufferManager10ITensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime13BufferManager10ITensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a5f8cccf8f635cf3cfbec353e1a7317b9"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">ITensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9UniquePtrE" title="tensorrt_llm::runtime::ITensor::UniquePtr"><span class="n"><span class="pre">UniquePtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10ITensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13BufferManager13CudaStreamPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13BufferManager13CudaStreamPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime13BufferManager13CudaStreamPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1ae977afd5a423417ac57f3bffa8162b33"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">CudaStreamPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10CudaStreamE" title="tensorrt_llm::runtime::CudaStream"><span class="n"><span class="pre">CudaStream</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager13CudaStreamPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13BufferManager14CudaMemPoolPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13BufferManager14CudaMemPoolPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime13BufferManager14CudaMemPoolPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a116255e34fbfe37cb623ef2e280eb9fc"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">CudaMemPoolPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">CudaMemPool</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager14CudaMemPoolPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13BufferManager13BufferManagerE13CudaStreamPtrb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13BufferManager13BufferManagerE13CudaStreamPtrb"></span><span id="_CPPv2N12tensorrt_llm7runtime13BufferManager13BufferManagerE13CudaStreamPtrb"></span><span id="tensorrt_llm::runtime::BufferManager::BufferManager__CudaStreamPtr.b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a1736c6a4c4a327a00a117addf7cfd915"></span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">BufferManager</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager13CudaStreamPtrE" title="tensorrt_llm::runtime::BufferManager::CudaStreamPtr"><span class="n"><span class="pre">CudaStreamPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">stream</span></span>, <span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">trimPool</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">false</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager13BufferManagerE13CudaStreamPtrb" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Construct a <a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1BufferManager"><span class="std std-ref">BufferManager</span></a>. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>cudaStream</strong> – <strong>[in]</strong> The cuda stream to use for all operations on GPU (allocation, de-allocation, copying, etc.). </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13BufferManagerD0Ev">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13BufferManagerD0Ev"></span><span id="_CPPv2N12tensorrt_llm7runtime13BufferManagerD0Ev"></span><span id="tensorrt_llm::runtime::BufferManager::~BufferManager"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1aea6ac13251be9f98419b4b20d9ee5b50"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">~BufferManager</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13BufferManagerD0Ev" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Destructor. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13BufferManager3gpuENSt6size_tEN8nvinfer18DataTypeE">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13BufferManager3gpuENSt6size_tEN8nvinfer18DataTypeE"></span><span id="_CPPv2NK12tensorrt_llm7runtime13BufferManager3gpuENSt6size_tEN8nvinfer18DataTypeE"></span><span id="tensorrt_llm::runtime::BufferManager::gpu__std::s.nvinfer1::DataTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a2cf0e00fd6ac0cfe8b593749583a7a03"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10IBufferPtrE" title="tensorrt_llm::runtime::BufferManager::IBufferPtr"><span class="n"><span class="pre">IBufferPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">gpu</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">type</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10kBYTE_TYPEE" title="tensorrt_llm::runtime::BufferManager::kBYTE_TYPE"><span class="n"><span class="pre">kBYTE_TYPE</span></span></a><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13BufferManager3gpuENSt6size_tEN8nvinfer18DataTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Allocates an <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1IBuffer"><span class="std std-ref"><span class="pre">IBuffer</span></span></a></code> of the given size on the GPU, using cudaMallocAsync. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13BufferManager3gpuEN8nvinfer14DimsEN8nvinfer18DataTypeE">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13BufferManager3gpuEN8nvinfer14DimsEN8nvinfer18DataTypeE"></span><span id="_CPPv2NK12tensorrt_llm7runtime13BufferManager3gpuEN8nvinfer14DimsEN8nvinfer18DataTypeE"></span><span id="tensorrt_llm::runtime::BufferManager::gpu__nvinfer1::Dims.nvinfer1::DataTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a5a023b2d5f629bc86e262f9733a12a59"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10ITensorPtrE" title="tensorrt_llm::runtime::BufferManager::ITensorPtr"><span class="n"><span class="pre">ITensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">gpu</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">Dims</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">dims</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">type</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10kBYTE_TYPEE" title="tensorrt_llm::runtime::BufferManager::kBYTE_TYPE"><span class="n"><span class="pre">kBYTE_TYPE</span></span></a><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13BufferManager3gpuEN8nvinfer14DimsEN8nvinfer18DataTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Allocates an <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1ITensor"><span class="std std-ref"><span class="pre">ITensor</span></span></a></code> of the given dimensions on the GPU, using cudaMallocAsync. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13BufferManager8allocateE10MemoryTypeNSt6size_tEN8nvinfer18DataTypeE">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13BufferManager8allocateE10MemoryTypeNSt6size_tEN8nvinfer18DataTypeE"></span><span id="_CPPv2NK12tensorrt_llm7runtime13BufferManager8allocateE10MemoryTypeNSt6size_tEN8nvinfer18DataTypeE"></span><span id="tensorrt_llm::runtime::BufferManager::allocate__MemoryType.std::s.nvinfer1::DataTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a1bdafec3fa633fdb0d321f5e748b3174"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10IBufferPtrE" title="tensorrt_llm::runtime::BufferManager::IBufferPtr"><span class="n"><span class="pre">IBufferPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">allocate</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="tensorrt_llm::runtime::MemoryType"><span class="n"><span class="pre">MemoryType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">memoryType</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">type</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10kBYTE_TYPEE" title="tensorrt_llm::runtime::BufferManager::kBYTE_TYPE"><span class="n"><span class="pre">kBYTE_TYPE</span></span></a><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13BufferManager8allocateE10MemoryTypeNSt6size_tEN8nvinfer18DataTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Allocates an <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1IBuffer"><span class="std std-ref"><span class="pre">IBuffer</span></span></a></code> of the given size and memory type. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13BufferManager8allocateE10MemoryTypeN8nvinfer14DimsEN8nvinfer18DataTypeE">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13BufferManager8allocateE10MemoryTypeN8nvinfer14DimsEN8nvinfer18DataTypeE"></span><span id="_CPPv2NK12tensorrt_llm7runtime13BufferManager8allocateE10MemoryTypeN8nvinfer14DimsEN8nvinfer18DataTypeE"></span><span id="tensorrt_llm::runtime::BufferManager::allocate__MemoryType.nvinfer1::Dims.nvinfer1::DataTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a0e4e4c0e248e95e19061a61fadec30cc"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10ITensorPtrE" title="tensorrt_llm::runtime::BufferManager::ITensorPtr"><span class="n"><span class="pre">ITensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">allocate</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="tensorrt_llm::runtime::MemoryType"><span class="n"><span class="pre">MemoryType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">memoryType</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">Dims</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">dims</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">type</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10kBYTE_TYPEE" title="tensorrt_llm::runtime::BufferManager::kBYTE_TYPE"><span class="n"><span class="pre">kBYTE_TYPE</span></span></a><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13BufferManager8allocateE10MemoryTypeN8nvinfer14DimsEN8nvinfer18DataTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Allocates an <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1ITensor"><span class="std std-ref"><span class="pre">ITensor</span></span></a></code> of the given dimensions and memory type. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13BufferManager11emptyBufferE10MemoryTypeN8nvinfer18DataTypeE">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13BufferManager11emptyBufferE10MemoryTypeN8nvinfer18DataTypeE"></span><span id="_CPPv2NK12tensorrt_llm7runtime13BufferManager11emptyBufferE10MemoryTypeN8nvinfer18DataTypeE"></span><span id="tensorrt_llm::runtime::BufferManager::emptyBuffer__MemoryType.nvinfer1::DataTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a0e78e3aa2aa1d635ae230996b449f9a4"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10IBufferPtrE" title="tensorrt_llm::runtime::BufferManager::IBufferPtr"><span class="n"><span class="pre">IBufferPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">emptyBuffer</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="tensorrt_llm::runtime::MemoryType"><span class="n"><span class="pre">MemoryType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">memoryType</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">type</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10kBYTE_TYPEE" title="tensorrt_llm::runtime::BufferManager::kBYTE_TYPE"><span class="n"><span class="pre">kBYTE_TYPE</span></span></a><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13BufferManager11emptyBufferE10MemoryTypeN8nvinfer18DataTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Create an empty <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1IBuffer"><span class="std std-ref"><span class="pre">IBuffer</span></span></a></code> of the given memory type. It may be resized later. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13BufferManager11emptyTensorE10MemoryTypeN8nvinfer18DataTypeE">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13BufferManager11emptyTensorE10MemoryTypeN8nvinfer18DataTypeE"></span><span id="_CPPv2NK12tensorrt_llm7runtime13BufferManager11emptyTensorE10MemoryTypeN8nvinfer18DataTypeE"></span><span id="tensorrt_llm::runtime::BufferManager::emptyTensor__MemoryType.nvinfer1::DataTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a1d0157fa19df03ba561d15d5b64de259"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10ITensorPtrE" title="tensorrt_llm::runtime::BufferManager::ITensorPtr"><span class="n"><span class="pre">ITensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">emptyTensor</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="tensorrt_llm::runtime::MemoryType"><span class="n"><span class="pre">MemoryType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">memoryType</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">type</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10kBYTE_TYPEE" title="tensorrt_llm::runtime::BufferManager::kBYTE_TYPE"><span class="n"><span class="pre">kBYTE_TYPE</span></span></a><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13BufferManager11emptyTensorE10MemoryTypeN8nvinfer18DataTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Create an empty <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1ITensor"><span class="std std-ref"><span class="pre">ITensor</span></span></a></code> of the given memory type. It may be reshaped later. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13BufferManager6setMemER7IBuffer7int32_t">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13BufferManager6setMemER7IBuffer7int32_t"></span><span id="_CPPv2NK12tensorrt_llm7runtime13BufferManager6setMemER7IBuffer7int32_t"></span><span id="tensorrt_llm::runtime::BufferManager::setMem__IBufferR.int32_tC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1ab625f396732dddcad22f44e9ab02fc28"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setMem</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">buffer</span></span>, <span class="n"><span class="pre">int32_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">value</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13BufferManager6setMemER7IBuffer7int32_t" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Set the contents of the given <code class="docutils literal notranslate"><span class="pre">buffer</span></code> to value. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13BufferManager7setZeroER7IBuffer">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13BufferManager7setZeroER7IBuffer"></span><span id="_CPPv2NK12tensorrt_llm7runtime13BufferManager7setZeroER7IBuffer"></span><span id="tensorrt_llm::runtime::BufferManager::setZero__IBufferRC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1afe66c31bfc8e00ebbdb8405194be0db9"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setZero</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">buffer</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13BufferManager7setZeroER7IBuffer" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Set the contents of the given <code class="docutils literal notranslate"><span class="pre">buffer</span></code> to zero. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13BufferManager4copyEPKvR7IBuffer10MemoryType">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13BufferManager4copyEPKvR7IBuffer10MemoryType"></span><span id="_CPPv2NK12tensorrt_llm7runtime13BufferManager4copyEPKvR7IBuffer10MemoryType"></span><span id="tensorrt_llm::runtime::BufferManager::copy__voidCP.IBufferR.MemoryTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a1adc12c1388e8a3392ac39a3ae64d518"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">copy</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="n sig-param"><span class="pre">src</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">dst</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="tensorrt_llm::runtime::MemoryType"><span class="n"><span class="pre">MemoryType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">srcType</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13BufferManager4copyEPKvR7IBuffer10MemoryType" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Copy <code class="docutils literal notranslate"><span class="pre">src</span></code> to <code class="docutils literal notranslate"><span class="pre">dst</span></code>. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13BufferManager4copyERK7IBufferPv10MemoryType">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13BufferManager4copyERK7IBufferPv10MemoryType"></span><span id="_CPPv2NK12tensorrt_llm7runtime13BufferManager4copyERK7IBufferPv10MemoryType"></span><span id="tensorrt_llm::runtime::BufferManager::copy__IBufferCR.voidP.MemoryTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a313f2bd39b26e568c97f092a98ac20f9"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">copy</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">src</span></span>, <span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="n sig-param"><span class="pre">dst</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="tensorrt_llm::runtime::MemoryType"><span class="n"><span class="pre">MemoryType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">dstType</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13BufferManager4copyERK7IBufferPv10MemoryType" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Copy <code class="docutils literal notranslate"><span class="pre">src</span></code> to <code class="docutils literal notranslate"><span class="pre">dst</span></code>. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13BufferManager4copyEPKvR7IBuffer">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13BufferManager4copyEPKvR7IBuffer"></span><span id="_CPPv2NK12tensorrt_llm7runtime13BufferManager4copyEPKvR7IBuffer"></span><span id="tensorrt_llm::runtime::BufferManager::copy__voidCP.IBufferRC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a3086524a384108c45199076f536d4d9c"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">copy</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="n sig-param"><span class="pre">src</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">dst</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13BufferManager4copyEPKvR7IBuffer" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Copy <code class="docutils literal notranslate"><span class="pre">src</span></code> to <code class="docutils literal notranslate"><span class="pre">dst</span></code>. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13BufferManager4copyERK7IBufferPv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13BufferManager4copyERK7IBufferPv"></span><span id="_CPPv2NK12tensorrt_llm7runtime13BufferManager4copyERK7IBufferPv"></span><span id="tensorrt_llm::runtime::BufferManager::copy__IBufferCR.voidPC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a583c0a02dcd51050b5b4273a19c0bde7"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">copy</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">src</span></span>, <span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="n sig-param"><span class="pre">dst</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13BufferManager4copyERK7IBufferPv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Copy <code class="docutils literal notranslate"><span class="pre">src</span></code> to <code class="docutils literal notranslate"><span class="pre">dst</span></code>. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13BufferManager4copyERK7IBufferR7IBuffer">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13BufferManager4copyERK7IBufferR7IBuffer"></span><span id="_CPPv2NK12tensorrt_llm7runtime13BufferManager4copyERK7IBufferR7IBuffer"></span><span id="tensorrt_llm::runtime::BufferManager::copy__IBufferCR.IBufferRC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a406d3f69a661d5a3bcb5dded51626c44"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">copy</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">src</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">dst</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13BufferManager4copyERK7IBufferR7IBuffer" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Copy <code class="docutils literal notranslate"><span class="pre">src</span></code> to <code class="docutils literal notranslate"><span class="pre">dst</span></code>. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13BufferManager8copyFromERK7IBuffer10MemoryType">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13BufferManager8copyFromERK7IBuffer10MemoryType"></span><span id="_CPPv2NK12tensorrt_llm7runtime13BufferManager8copyFromERK7IBuffer10MemoryType"></span><span id="tensorrt_llm::runtime::BufferManager::copyFrom__IBufferCR.MemoryTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a3101d71e18fa84154b687236487dfc5f"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10IBufferPtrE" title="tensorrt_llm::runtime::BufferManager::IBufferPtr"><span class="n"><span class="pre">IBufferPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">copyFrom</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">src</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="tensorrt_llm::runtime::MemoryType"><span class="n"><span class="pre">MemoryType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">memoryType</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13BufferManager8copyFromERK7IBuffer10MemoryType" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Copy <code class="docutils literal notranslate"><span class="pre">src</span></code> into a new <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1IBuffer"><span class="std std-ref"><span class="pre">IBuffer</span></span></a></code> with a potentially different memory type. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13BufferManager8copyFromERK7ITensor10MemoryType">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13BufferManager8copyFromERK7ITensor10MemoryType"></span><span id="_CPPv2NK12tensorrt_llm7runtime13BufferManager8copyFromERK7ITensor10MemoryType"></span><span id="tensorrt_llm::runtime::BufferManager::copyFrom__ITensorCR.MemoryTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a80a336bdfbd0faf2d76e470354bcddf0"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10ITensorPtrE" title="tensorrt_llm::runtime::BufferManager::ITensorPtr"><span class="n"><span class="pre">ITensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">copyFrom</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">src</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="tensorrt_llm::runtime::MemoryType"><span class="n"><span class="pre">MemoryType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">memoryType</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13BufferManager8copyFromERK7ITensor10MemoryType" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Copy <code class="docutils literal notranslate"><span class="pre">src</span></code> into a new <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1ITensor"><span class="std std-ref"><span class="pre">ITensor</span></span></a></code> with a potentially different memory type. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0ENK12tensorrt_llm7runtime13BufferManager8copyFromE10IBufferPtrRKNSt6vectorI1TEE10MemoryType">
|
||
<span id="_CPPv3I0ENK12tensorrt_llm7runtime13BufferManager8copyFromERKNSt6vectorI1TEE10MemoryType"></span><span id="_CPPv2I0ENK12tensorrt_llm7runtime13BufferManager8copyFromERKNSt6vectorI1TEE10MemoryType"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a51f44f1cf97f9c0b686048d37291ea5a"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10IBufferPtrE" title="tensorrt_llm::runtime::BufferManager::IBufferPtr"><span class="n"><span class="pre">IBufferPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">copyFrom</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0ENK12tensorrt_llm7runtime13BufferManager8copyFromE10IBufferPtrRKNSt6vectorI1TEE10MemoryType" title="tensorrt_llm::runtime::BufferManager::copyFrom::T"><span class="n"><span class="pre">T</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">src</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="tensorrt_llm::runtime::MemoryType"><span class="n"><span class="pre">MemoryType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">memoryType</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4I0ENK12tensorrt_llm7runtime13BufferManager8copyFromE10IBufferPtrRKNSt6vectorI1TEE10MemoryType" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Copy <code class="docutils literal notranslate"><span class="pre">src</span></code> into a new <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1IBuffer"><span class="std std-ref"><span class="pre">IBuffer</span></span></a></code> with a potentially different memory type. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0ENK12tensorrt_llm7runtime13BufferManager8copyFromE10ITensorPtrP1TN8nvinfer14DimsE10MemoryType">
|
||
<span id="_CPPv3I0ENK12tensorrt_llm7runtime13BufferManager8copyFromEP1TN8nvinfer14DimsE10MemoryType"></span><span id="_CPPv2I0ENK12tensorrt_llm7runtime13BufferManager8copyFromEP1TN8nvinfer14DimsE10MemoryType"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1ac42305510a739c47d3c4358b97560f3e"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10ITensorPtrE" title="tensorrt_llm::runtime::BufferManager::ITensorPtr"><span class="n"><span class="pre">ITensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">copyFrom</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4I0ENK12tensorrt_llm7runtime13BufferManager8copyFromE10ITensorPtrP1TN8nvinfer14DimsE10MemoryType" title="tensorrt_llm::runtime::BufferManager::copyFrom::T"><span class="n"><span class="pre">T</span></span></a><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="n sig-param"><span class="pre">src</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">Dims</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">dims</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="tensorrt_llm::runtime::MemoryType"><span class="n"><span class="pre">MemoryType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">memoryType</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4I0ENK12tensorrt_llm7runtime13BufferManager8copyFromE10ITensorPtrP1TN8nvinfer14DimsE10MemoryType" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Copy <code class="docutils literal notranslate"><span class="pre">src</span></code> into a new <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1ITensor"><span class="std std-ref"><span class="pre">ITensor</span></span></a></code> with a potentially different memory type. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0ENK12tensorrt_llm7runtime13BufferManager8copyFromE10ITensorPtrRKNSt6vectorI1TEEN8nvinfer14DimsE10MemoryType">
|
||
<span id="_CPPv3I0ENK12tensorrt_llm7runtime13BufferManager8copyFromERKNSt6vectorI1TEEN8nvinfer14DimsE10MemoryType"></span><span id="_CPPv2I0ENK12tensorrt_llm7runtime13BufferManager8copyFromERKNSt6vectorI1TEEN8nvinfer14DimsE10MemoryType"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a8feac5727a72ee0d1c50b3ff5ee4e979"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10ITensorPtrE" title="tensorrt_llm::runtime::BufferManager::ITensorPtr"><span class="n"><span class="pre">ITensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">copyFrom</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0ENK12tensorrt_llm7runtime13BufferManager8copyFromE10ITensorPtrRKNSt6vectorI1TEEN8nvinfer14DimsE10MemoryType" title="tensorrt_llm::runtime::BufferManager::copyFrom::T"><span class="n"><span class="pre">T</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">src</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">Dims</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">dims</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="tensorrt_llm::runtime::MemoryType"><span class="n"><span class="pre">MemoryType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">memoryType</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4I0ENK12tensorrt_llm7runtime13BufferManager8copyFromE10ITensorPtrRKNSt6vectorI1TEEN8nvinfer14DimsE10MemoryType" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Copy <code class="docutils literal notranslate"><span class="pre">src</span></code> into a new <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1ITensor"><span class="std std-ref"><span class="pre">ITensor</span></span></a></code> with a potentially different memory type. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13BufferManager9getStreamEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13BufferManager9getStreamEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime13BufferManager9getStreamEv"></span><span id="tensorrt_llm::runtime::BufferManager::getStreamC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a895d88127b8bc55f374727908284e352"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10CudaStreamE" title="tensorrt_llm::runtime::CudaStream"><span class="n"><span class="pre">CudaStream</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="n"><span class="pre">getStream</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13BufferManager9getStreamEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Get the underlying cuda stream. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13BufferManager18memoryPoolReservedEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13BufferManager18memoryPoolReservedEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime13BufferManager18memoryPoolReservedEv"></span><span id="tensorrt_llm::runtime::BufferManager::memoryPoolReservedC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a92a21be0a954deb04c9c6e8540a2141b"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">memoryPoolReserved</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13BufferManager18memoryPoolReservedEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>The current size of the memory reserved by the memory pool. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13BufferManager14memoryPoolUsedEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13BufferManager14memoryPoolUsedEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime13BufferManager14memoryPoolUsedEv"></span><span id="tensorrt_llm::runtime::BufferManager::memoryPoolUsedC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a73ae2808af584ba85c1a61134f025659"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">memoryPoolUsed</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13BufferManager14memoryPoolUsedEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>The current size of the memory used by the memory pool. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13BufferManager14memoryPoolFreeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13BufferManager14memoryPoolFreeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime13BufferManager14memoryPoolFreeEv"></span><span id="tensorrt_llm::runtime::BufferManager::memoryPoolFreeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a822c7eaa19d0bdebd460fdc4cf6ee82a"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">memoryPoolFree</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13BufferManager14memoryPoolFreeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>The current size of the memory free in the memory pool. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13BufferManager16memoryPoolTrimToENSt6size_tE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13BufferManager16memoryPoolTrimToENSt6size_tE"></span><span id="_CPPv2N12tensorrt_llm7runtime13BufferManager16memoryPoolTrimToENSt6size_tE"></span><span id="tensorrt_llm::runtime::BufferManager::memoryPoolTrimTo__std::s"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a5e74ccf0e2e2132ae5834c31209f87d3"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">memoryPoolTrimTo</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager16memoryPoolTrimToENSt6size_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Try to trim the memory reserved by the pool to <code class="docutils literal notranslate"><span class="pre">size</span></code> bytes. This synchronizes implicitly with the stream. </p>
|
||
</dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-functions">Public Static Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13BufferManager7gpuSyncENSt6size_tEN8nvinfer18DataTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13BufferManager7gpuSyncENSt6size_tEN8nvinfer18DataTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime13BufferManager7gpuSyncENSt6size_tEN8nvinfer18DataTypeE"></span><span id="tensorrt_llm::runtime::BufferManager::gpuSync__std::s.nvinfer1::DataType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a5906b574a0a12f5842f4e4743fad0d02"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10IBufferPtrE" title="tensorrt_llm::runtime::BufferManager::IBufferPtr"><span class="n"><span class="pre">IBufferPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">gpuSync</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">type</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10kBYTE_TYPEE" title="tensorrt_llm::runtime::BufferManager::kBYTE_TYPE"><span class="n"><span class="pre">kBYTE_TYPE</span></span></a><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager7gpuSyncENSt6size_tEN8nvinfer18DataTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Allocates an <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1IBuffer"><span class="std std-ref"><span class="pre">IBuffer</span></span></a></code> of the given size on the GPU, using cudaMalloc. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13BufferManager7gpuSyncEN8nvinfer14DimsEN8nvinfer18DataTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13BufferManager7gpuSyncEN8nvinfer14DimsEN8nvinfer18DataTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime13BufferManager7gpuSyncEN8nvinfer14DimsEN8nvinfer18DataTypeE"></span><span id="tensorrt_llm::runtime::BufferManager::gpuSync__nvinfer1::Dims.nvinfer1::DataType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a0cd267d74d606d1b2f19c64d6dacab9a"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10ITensorPtrE" title="tensorrt_llm::runtime::BufferManager::ITensorPtr"><span class="n"><span class="pre">ITensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">gpuSync</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">Dims</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">dims</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">type</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10kBYTE_TYPEE" title="tensorrt_llm::runtime::BufferManager::kBYTE_TYPE"><span class="n"><span class="pre">kBYTE_TYPE</span></span></a><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager7gpuSyncEN8nvinfer14DimsEN8nvinfer18DataTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Allocates an <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1ITensor"><span class="std std-ref"><span class="pre">ITensor</span></span></a></code> of the given dimensions on the GPU, using cudaMalloc. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13BufferManager3cpuENSt6size_tEN8nvinfer18DataTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13BufferManager3cpuENSt6size_tEN8nvinfer18DataTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime13BufferManager3cpuENSt6size_tEN8nvinfer18DataTypeE"></span><span id="tensorrt_llm::runtime::BufferManager::cpu__std::s.nvinfer1::DataType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a7edb3cdde9b85c01fec18f3f4f34282f"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10IBufferPtrE" title="tensorrt_llm::runtime::BufferManager::IBufferPtr"><span class="n"><span class="pre">IBufferPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">cpu</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">type</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10kBYTE_TYPEE" title="tensorrt_llm::runtime::BufferManager::kBYTE_TYPE"><span class="n"><span class="pre">kBYTE_TYPE</span></span></a><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager3cpuENSt6size_tEN8nvinfer18DataTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Allocates an <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1IBuffer"><span class="std std-ref"><span class="pre">IBuffer</span></span></a></code> of the given size on the CPU. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13BufferManager3cpuEN8nvinfer14DimsEN8nvinfer18DataTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13BufferManager3cpuEN8nvinfer14DimsEN8nvinfer18DataTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime13BufferManager3cpuEN8nvinfer14DimsEN8nvinfer18DataTypeE"></span><span id="tensorrt_llm::runtime::BufferManager::cpu__nvinfer1::Dims.nvinfer1::DataType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a3beb40d9ddabd96774167df148efc676"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10ITensorPtrE" title="tensorrt_llm::runtime::BufferManager::ITensorPtr"><span class="n"><span class="pre">ITensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">cpu</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">Dims</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">dims</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">type</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10kBYTE_TYPEE" title="tensorrt_llm::runtime::BufferManager::kBYTE_TYPE"><span class="n"><span class="pre">kBYTE_TYPE</span></span></a><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager3cpuEN8nvinfer14DimsEN8nvinfer18DataTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Allocates an <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1ITensor"><span class="std std-ref"><span class="pre">ITensor</span></span></a></code> of the given dimensions on the CPU. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13BufferManager6pinnedENSt6size_tEN8nvinfer18DataTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13BufferManager6pinnedENSt6size_tEN8nvinfer18DataTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime13BufferManager6pinnedENSt6size_tEN8nvinfer18DataTypeE"></span><span id="tensorrt_llm::runtime::BufferManager::pinned__std::s.nvinfer1::DataType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a6a58588cfc27c05c1d7dbb2f043672d3"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10IBufferPtrE" title="tensorrt_llm::runtime::BufferManager::IBufferPtr"><span class="n"><span class="pre">IBufferPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">pinned</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">type</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10kBYTE_TYPEE" title="tensorrt_llm::runtime::BufferManager::kBYTE_TYPE"><span class="n"><span class="pre">kBYTE_TYPE</span></span></a><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager6pinnedENSt6size_tEN8nvinfer18DataTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Allocates a pinned <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1IBuffer"><span class="std std-ref"><span class="pre">IBuffer</span></span></a></code> of the given size on the CPU. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13BufferManager6pinnedEN8nvinfer14DimsEN8nvinfer18DataTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13BufferManager6pinnedEN8nvinfer14DimsEN8nvinfer18DataTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime13BufferManager6pinnedEN8nvinfer14DimsEN8nvinfer18DataTypeE"></span><span id="tensorrt_llm::runtime::BufferManager::pinned__nvinfer1::Dims.nvinfer1::DataType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a3cfc4376c5e62876586e1e07e08d3069"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10ITensorPtrE" title="tensorrt_llm::runtime::BufferManager::ITensorPtr"><span class="n"><span class="pre">ITensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">pinned</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">Dims</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">dims</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">type</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10kBYTE_TYPEE" title="tensorrt_llm::runtime::BufferManager::kBYTE_TYPE"><span class="n"><span class="pre">kBYTE_TYPE</span></span></a><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager6pinnedEN8nvinfer14DimsEN8nvinfer18DataTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Allocates a pinned <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1ITensor"><span class="std std-ref"><span class="pre">ITensor</span></span></a></code> of the given dimensions on the CPU. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13BufferManager10pinnedPoolENSt6size_tEN8nvinfer18DataTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13BufferManager10pinnedPoolENSt6size_tEN8nvinfer18DataTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime13BufferManager10pinnedPoolENSt6size_tEN8nvinfer18DataTypeE"></span><span id="tensorrt_llm::runtime::BufferManager::pinnedPool__std::s.nvinfer1::DataType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a2e2ddfd2b16255d06ec4490a3bb04301"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10IBufferPtrE" title="tensorrt_llm::runtime::BufferManager::IBufferPtr"><span class="n"><span class="pre">IBufferPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">pinnedPool</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">type</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10kBYTE_TYPEE" title="tensorrt_llm::runtime::BufferManager::kBYTE_TYPE"><span class="n"><span class="pre">kBYTE_TYPE</span></span></a><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10pinnedPoolENSt6size_tEN8nvinfer18DataTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Allocates a pinned <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1IBuffer"><span class="std std-ref"><span class="pre">IBuffer</span></span></a></code> of the given size on the CPU in the default memory pool. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13BufferManager10pinnedPoolEN8nvinfer14DimsEN8nvinfer18DataTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13BufferManager10pinnedPoolEN8nvinfer14DimsEN8nvinfer18DataTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime13BufferManager10pinnedPoolEN8nvinfer14DimsEN8nvinfer18DataTypeE"></span><span id="tensorrt_llm::runtime::BufferManager::pinnedPool__nvinfer1::Dims.nvinfer1::DataType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a7a27f3c16ba4d2e85fe825170a7f2ed7"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10ITensorPtrE" title="tensorrt_llm::runtime::BufferManager::ITensorPtr"><span class="n"><span class="pre">ITensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">pinnedPool</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">Dims</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">dims</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">type</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10kBYTE_TYPEE" title="tensorrt_llm::runtime::BufferManager::kBYTE_TYPE"><span class="n"><span class="pre">kBYTE_TYPE</span></span></a><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10pinnedPoolEN8nvinfer14DimsEN8nvinfer18DataTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Allocates a pinned <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1ITensor"><span class="std std-ref"><span class="pre">ITensor</span></span></a></code> of the given dimensions on the CPU in the default memory pool. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13BufferManager7managedENSt6size_tEN8nvinfer18DataTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13BufferManager7managedENSt6size_tEN8nvinfer18DataTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime13BufferManager7managedENSt6size_tEN8nvinfer18DataTypeE"></span><span id="tensorrt_llm::runtime::BufferManager::managed__std::s.nvinfer1::DataType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1ace6b7b5bbcd7163c81afdbd7458e1124"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10IBufferPtrE" title="tensorrt_llm::runtime::BufferManager::IBufferPtr"><span class="n"><span class="pre">IBufferPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">managed</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">type</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10kBYTE_TYPEE" title="tensorrt_llm::runtime::BufferManager::kBYTE_TYPE"><span class="n"><span class="pre">kBYTE_TYPE</span></span></a><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager7managedENSt6size_tEN8nvinfer18DataTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Allocates an <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1IBuffer"><span class="std std-ref"><span class="pre">IBuffer</span></span></a></code> of the given size in UVM. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13BufferManager7managedEN8nvinfer14DimsEN8nvinfer18DataTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13BufferManager7managedEN8nvinfer14DimsEN8nvinfer18DataTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime13BufferManager7managedEN8nvinfer14DimsEN8nvinfer18DataTypeE"></span><span id="tensorrt_llm::runtime::BufferManager::managed__nvinfer1::Dims.nvinfer1::DataType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a101c32638ce8ad5ff06337e322c4bbe9"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10ITensorPtrE" title="tensorrt_llm::runtime::BufferManager::ITensorPtr"><span class="n"><span class="pre">ITensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">managed</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">Dims</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">dims</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">type</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10kBYTE_TYPEE" title="tensorrt_llm::runtime::BufferManager::kBYTE_TYPE"><span class="n"><span class="pre">kBYTE_TYPE</span></span></a><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager7managedEN8nvinfer14DimsEN8nvinfer18DataTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Allocates an <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1ITensor"><span class="std std-ref"><span class="pre">ITensor</span></span></a></code> of the given dimensions in UVM. </p>
|
||
</dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13BufferManager10kBYTE_TYPEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13BufferManager10kBYTE_TYPEE"></span><span id="_CPPv2N12tensorrt_llm7runtime13BufferManager10kBYTE_TYPEE"></span><span id="tensorrt_llm::runtime::BufferManager::kBYTE_TYPE__auto"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a154196e71e943c8f4c6d0ccacb2dfe3c"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kBYTE_TYPE</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">kUINT8</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager10kBYTE_TYPEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-members">Private Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13BufferManager7mStreamE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13BufferManager7mStreamE"></span><span id="_CPPv2N12tensorrt_llm7runtime13BufferManager7mStreamE"></span><span id="tensorrt_llm::runtime::BufferManager::mStream__CudaStreamPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1aeab688938251784e88b66cff01a0dde3"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager13CudaStreamPtrE" title="tensorrt_llm::runtime::BufferManager::CudaStreamPtr"><span class="n"><span class="pre">CudaStreamPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mStream</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager7mStreamE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13BufferManager5mPoolE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13BufferManager5mPoolE"></span><span id="_CPPv2N12tensorrt_llm7runtime13BufferManager5mPoolE"></span><span id="tensorrt_llm::runtime::BufferManager::mPool__CudaMemPoolPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a1e8ae1268c1ef21cabad78420b798e48"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager14CudaMemPoolPtrE" title="tensorrt_llm::runtime::BufferManager::CudaMemPoolPtr"><span class="n"><span class="pre">CudaMemPoolPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mPool</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager5mPoolE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13BufferManager9mTrimPoolE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13BufferManager9mTrimPoolE"></span><span id="_CPPv2N12tensorrt_llm7runtime13BufferManager9mTrimPoolE"></span><span id="tensorrt_llm::runtime::BufferManager::mTrimPool__bC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1a9e8da67a80a77356d5e3c14749a263a0"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mTrimPool</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager9mTrimPoolE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-friends">Friends</p>
|
||
<dl>
|
||
<dt class="sig sig-object cpp">
|
||
<em class="property"><span class="pre">friend</span> <span class="pre">class</span></em> <span class="pre">::BufferManagerTest</span></dt>
|
||
</dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="common-h">
|
||
<h2>common.h<a class="headerlink" href="#common-h" title="Link to this heading"></a></h2>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-defines">Defines</p>
|
||
<dl class="cpp macro">
|
||
<dt class="sig sig-object cpp" id="c.FMT_DIM">
|
||
<span class="target" id="common_8h_1a510c0e5d6315b189e4726c3dd6a76271"></span><span class="sig-name descname"><span class="n"><span class="pre">FMT_DIM</span></span></span><a class="headerlink" href="#c.FMT_DIM" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-typedefs">Typedefs</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10SizeType32E">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10SizeType32E"></span><span id="_CPPv2N12tensorrt_llm7runtime10SizeType32E"></span><span class="target" id="common_8h_1a3bdb407122ee5c03962b4aea2f6c61e8"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">SizeType32</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">int32_t</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10SizeType64E">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10SizeType64E"></span><span id="_CPPv2N12tensorrt_llm7runtime10SizeType64E"></span><span class="target" id="common_8h_1aca6041db22beb41ce3c4640c45058773"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">SizeType64</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">int64_t</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10SizeType64E" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11TokenIdTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11TokenIdTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11TokenIdTypeE"></span><span class="target" id="common_8h_1a1de916f1c3f3c8d1f9c66320afc6df17"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TokenIdType</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">int32_t</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11TokenIdTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14LoraTaskIdTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14LoraTaskIdTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14LoraTaskIdTypeE"></span><span class="target" id="common_8h_1aa4d6a559b4a19f8fbab65e8e7a0e69fe"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">LoraTaskIdType</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">uint64_t</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14LoraTaskIdTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime16TokenExtraIdTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16TokenExtraIdTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime16TokenExtraIdTypeE"></span><span class="target" id="common_8h_1a6fbec83529ee7fceb176b465d97f5d6e"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TokenExtraIdType</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">uint64_t</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime16TokenExtraIdTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime16VecTokenExtraIdsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16VecTokenExtraIdsE"></span><span id="_CPPv2N12tensorrt_llm7runtime16VecTokenExtraIdsE"></span><span class="target" id="common_8h_1a4df1b36fecce49a24d250a14ae2b7d85"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">VecTokenExtraIds</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16TokenExtraIdTypeE" title="tensorrt_llm::runtime::TokenExtraIdType"><span class="n"><span class="pre">TokenExtraIdType</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime16VecTokenExtraIdsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15VecUniqueTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15VecUniqueTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime15VecUniqueTokensE"></span><span class="target" id="common_8h_1a119cacfef2e257e99f248ee75116134c"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">VecUniqueTokens</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11UniqueTokenE" title="tensorrt_llm::runtime::UniqueToken"><span class="n"><span class="pre">UniqueToken</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15VecUniqueTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime12StringPtrMapE">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime12StringPtrMapE"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime12StringPtrMapE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="common_8h_1a43946c471b82feb36a6350de9cde277d"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">StringPtrMap</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">unordered_map</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime12StringPtrMapE" title="tensorrt_llm::runtime::StringPtrMap::T"><span class="n"><span class="pre">T</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime12StringPtrMapE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-enums">Enums</p>
|
||
<dl class="cpp enum-class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11RequestTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11RequestTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11RequestTypeE"></span><span class="target" id="common_8h_1a919e1e7b6860dffaef8f53660c992ca0"></span><span class="k"><span class="pre">enum</span></span><span class="w"> </span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">RequestType</span></span></span><span class="w"> </span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">int32_t</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11RequestTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p><em>Values:</em></p>
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11RequestType8kCONTEXTE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11RequestType8kCONTEXTE"></span><span id="_CPPv2N12tensorrt_llm7runtime11RequestType8kCONTEXTE"></span><span class="target" id="common_8h_1a919e1e7b6860dffaef8f53660c992ca0ab313e6f758ff978c83b115d402efb5f0"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kCONTEXT</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11RequestType8kCONTEXTE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11RequestType11kGENERATIONE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11RequestType11kGENERATIONE"></span><span id="_CPPv2N12tensorrt_llm7runtime11RequestType11kGENERATIONE"></span><span class="target" id="common_8h_1a919e1e7b6860dffaef8f53660c992ca0ad3fecc7bf972e65c8bc64551251be711"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kGENERATION</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11RequestType11kGENERATIONE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</div>
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11UniqueTokenE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11UniqueTokenE"></span><span id="_CPPv2N12tensorrt_llm7runtime11UniqueTokenE"></span><span id="tensorrt_llm::runtime::UniqueToken"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1UniqueToken"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">UniqueToken</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11UniqueTokenE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11UniqueTokeneqERK11UniqueToken">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11UniqueTokeneqERK11UniqueToken"></span><span id="_CPPv2NK12tensorrt_llm7runtime11UniqueTokeneqERK11UniqueToken"></span><span id="tensorrt_llm::runtime::UniqueToken::eq-operator__UniqueTokenCRC"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1UniqueToken_1ac13480743f9ec9bb14da311d96ed9536"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="k"><span class="pre">operator</span></span><span class="o"><span class="pre">==</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11UniqueTokenE" title="tensorrt_llm::runtime::UniqueToken"><span class="n"><span class="pre">UniqueToken</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">other</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11UniqueTokeneqERK11UniqueToken" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11UniqueToken7tokenIdE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11UniqueToken7tokenIdE"></span><span id="_CPPv2N12tensorrt_llm7runtime11UniqueToken7tokenIdE"></span><span id="tensorrt_llm::runtime::UniqueToken::tokenId__TokenIdType"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1UniqueToken_1a126d61de5902884d3a08a85a502b4afc"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11TokenIdTypeE" title="tensorrt_llm::runtime::TokenIdType"><span class="n"><span class="pre">TokenIdType</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tokenId</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11UniqueToken7tokenIdE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11UniqueToken12tokenExtraIdE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11UniqueToken12tokenExtraIdE"></span><span id="_CPPv2N12tensorrt_llm7runtime11UniqueToken12tokenExtraIdE"></span><span id="tensorrt_llm::runtime::UniqueToken::tokenExtraId__TokenExtraIdType"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1UniqueToken_1af5dc28b61aa34a4ae8a01f85695bfdd3"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16TokenExtraIdTypeE" title="tensorrt_llm::runtime::TokenExtraIdType"><span class="n"><span class="pre">TokenExtraIdType</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tokenExtraId</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11UniqueToken12tokenExtraIdE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="cudaevent-h">
|
||
<h2>cudaEvent.h<a class="headerlink" href="#cudaevent-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9CudaEventE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9CudaEventE"></span><span id="_CPPv2N12tensorrt_llm7runtime9CudaEventE"></span><span id="tensorrt_llm::runtime::CudaEvent"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaEvent"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">CudaEvent</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9CudaEventE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9CudaEvent7pointerE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9CudaEvent7pointerE"></span><span id="_CPPv2N12tensorrt_llm7runtime9CudaEvent7pointerE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaEvent_1a765323cdb24844d31c8f9e354a5194f8"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">pointer</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">cudaEvent_t</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9CudaEvent7pointerE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9CudaEvent9CudaEventEj">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9CudaEvent9CudaEventEj"></span><span id="_CPPv2N12tensorrt_llm7runtime9CudaEvent9CudaEventEj"></span><span id="tensorrt_llm::runtime::CudaEvent::CudaEvent__unsigned-i"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaEvent_1aae9d65fd3cc3d42763c1219710dcd6ac"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">CudaEvent</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">unsigned</span></span><span class="w"> </span><span class="kt"><span class="pre">int</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">flags</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">cudaEventDisableTiming</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9CudaEvent9CudaEventEj" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Creates a new cuda event. The event will be destroyed in the destructor.</p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>flags</strong> – Flags for event creation. By default, event timing is disabled. </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9CudaEvent9CudaEventE7pointerb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9CudaEvent9CudaEventE7pointerb"></span><span id="_CPPv2N12tensorrt_llm7runtime9CudaEvent9CudaEventE7pointerb"></span><span id="tensorrt_llm::runtime::CudaEvent::CudaEvent__pointer.b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaEvent_1a8ac4cb3f5ac924e72862c1c5fd033cbd"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">CudaEvent</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEvent7pointerE" title="tensorrt_llm::runtime::CudaEvent::pointer"><span class="n"><span class="pre">pointer</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">event</span></span>, <span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">ownsEvent</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">true</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9CudaEvent9CudaEventE7pointerb" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Pass an existing cuda event to this object.</p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><ul class="simple">
|
||
<li><p><strong>event</strong> – The event to pass to this object. </p></li>
|
||
<li><p><strong>ownsEvent</strong> – Whether this object owns the event and destroys it in the destructor. </p></li>
|
||
</ul>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime9CudaEvent3getEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime9CudaEvent3getEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime9CudaEvent3getEv"></span><span id="tensorrt_llm::runtime::CudaEvent::getC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaEvent_1a530c756fe9e6ad149b813659b3644f16"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEvent7pointerE" title="tensorrt_llm::runtime::CudaEvent::pointer"><span class="n"><span class="pre">pointer</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">get</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime9CudaEvent3getEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Returns the event associated with this object. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime9CudaEvent11synchronizeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime9CudaEvent11synchronizeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime9CudaEvent11synchronizeEv"></span><span id="tensorrt_llm::runtime::CudaEvent::synchronizeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaEvent_1a9216cdfafea99849fb47b1be60fb1a7e"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">synchronize</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime9CudaEvent11synchronizeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Synchronizes the event. </p>
|
||
</dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-types">Private Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9CudaEvent12element_typeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9CudaEvent12element_typeE"></span><span id="_CPPv2N12tensorrt_llm7runtime9CudaEvent12element_typeE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaEvent_1a411326cde2f9f947c5d92abe724ebb34"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">element_type</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">remove_pointer_t</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEvent7pointerE" title="tensorrt_llm::runtime::CudaEvent::pointer"><span class="n"><span class="pre">pointer</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9CudaEvent12element_typeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9CudaEvent8EventPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9CudaEvent8EventPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime9CudaEvent8EventPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaEvent_1ac5468fc310bab51276ed6d6212e7240f"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">EventPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">unique_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEvent12element_typeE" title="tensorrt_llm::runtime::CudaEvent::element_type"><span class="n"><span class="pre">element_type</span></span></a><span class="p"><span class="pre">,</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEvent7DeleterE" title="tensorrt_llm::runtime::CudaEvent::Deleter"><span class="n"><span class="pre">Deleter</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9CudaEvent8EventPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-members">Private Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9CudaEvent6mEventE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9CudaEvent6mEventE"></span><span id="_CPPv2N12tensorrt_llm7runtime9CudaEvent6mEventE"></span><span id="tensorrt_llm::runtime::CudaEvent::mEvent__EventPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaEvent_1a3231de6317de1857e5f699a94d59dcf2"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEvent8EventPtrE" title="tensorrt_llm::runtime::CudaEvent::EventPtr"><span class="n"><span class="pre">EventPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mEvent</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9CudaEvent6mEventE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9CudaEvent7DeleterE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9CudaEvent7DeleterE"></span><span id="_CPPv2N12tensorrt_llm7runtime9CudaEvent7DeleterE"></span><span id="tensorrt_llm::runtime::CudaEvent::Deleter"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaEvent_1_1Deleter"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Deleter</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9CudaEvent7DeleterE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9CudaEvent7Deleter7DeleterEb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9CudaEvent7Deleter7DeleterEb"></span><span id="_CPPv2N12tensorrt_llm7runtime9CudaEvent7Deleter7DeleterEb"></span><span id="tensorrt_llm::runtime::CudaEvent::Deleter::Deleter__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaEvent_1_1Deleter_1a7807c6a0240c9d01f09dbde19b422d05"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Deleter</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">ownsEvent</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9CudaEvent7Deleter7DeleterEb" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9CudaEvent7Deleter7DeleterEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9CudaEvent7Deleter7DeleterEv"></span><span id="_CPPv2N12tensorrt_llm7runtime9CudaEvent7Deleter7DeleterEv"></span><span id="tensorrt_llm::runtime::CudaEvent::Deleter::Deleter"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaEvent_1_1Deleter_1aec3a7bdb695dc6598d6c6b8c5cbd4598"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Deleter</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9CudaEvent7Deleter7DeleterEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime9CudaEvent7DeleterclE7pointer">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime9CudaEvent7DeleterclE7pointer"></span><span id="_CPPv2NK12tensorrt_llm7runtime9CudaEvent7DeleterclE7pointer"></span><span id="tensorrt_llm::runtime::CudaEvent::Deleter::call-operator__pointerCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaEvent_1_1Deleter_1a82af09350919fceb1f0f4c43d1822d37"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="k"><span class="pre">operator</span></span><span class="o"><span class="pre">()</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEvent7pointerE" title="tensorrt_llm::runtime::CudaEvent::pointer"><span class="n"><span class="pre">pointer</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">event</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime9CudaEvent7DeleterclE7pointer" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-members">Private Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9CudaEvent7Deleter10mOwnsEventE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9CudaEvent7Deleter10mOwnsEventE"></span><span id="_CPPv2N12tensorrt_llm7runtime9CudaEvent7Deleter10mOwnsEventE"></span><span id="tensorrt_llm::runtime::CudaEvent::Deleter::mOwnsEvent__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaEvent_1_1Deleter_1a4df7640a97930eaf43a1ca952f3f919f"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mOwnsEvent</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9CudaEvent7Deleter10mOwnsEventE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="cudastream-h">
|
||
<h2>cudaStream.h<a class="headerlink" href="#cudastream-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10CudaStreamE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10CudaStreamE"></span><span id="_CPPv2N12tensorrt_llm7runtime10CudaStreamE"></span><span id="tensorrt_llm::runtime::CudaStream"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaStream"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">CudaStream</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10CudaStreamE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10CudaStream10CudaStreamEji">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10CudaStream10CudaStreamEji"></span><span id="_CPPv2N12tensorrt_llm7runtime10CudaStream10CudaStreamEji"></span><span id="tensorrt_llm::runtime::CudaStream::CudaStream__unsigned-i.i"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaStream_1a2ba13a61587813c68c018a64ed2967fe"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">CudaStream</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">unsigned</span></span><span class="w"> </span><span class="kt"><span class="pre">int</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">flags</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">cudaStreamNonBlocking</span></span>, <span class="kt"><span class="pre">int</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">priority</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10CudaStream10CudaStreamEji" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Creates a new cuda stream on the current device. The stream will be destroyed in the destructor.</p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><ul class="simple">
|
||
<li><p><strong>flags</strong> – Flags for stream creation. See ::cudaStreamCreateWithFlags for a list of valid flags that can be passed. </p></li>
|
||
<li><p><strong>priority</strong> – Priority of the stream. Lower numbers represent higher priorities. See ::cudaDeviceGetStreamPriorityRange for more information about the meaningful stream priorities that can be passed. </p></li>
|
||
</ul>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10CudaStream10CudaStreamE12cudaStream_tib">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10CudaStream10CudaStreamE12cudaStream_tib"></span><span id="_CPPv2N12tensorrt_llm7runtime10CudaStream10CudaStreamE12cudaStream_tib"></span><span id="tensorrt_llm::runtime::CudaStream::CudaStream__cudaStream_t.i.b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaStream_1a4ff09ea22fc6679e2d93b772e148d19e"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">CudaStream</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">cudaStream_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">stream</span></span>, <span class="kt"><span class="pre">int</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">device</span></span>, <span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">ownsStream</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">true</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10CudaStream10CudaStreamE12cudaStream_tib" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Pass an existing cuda stream to this object.</p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><ul class="simple">
|
||
<li><p><strong>stream</strong> – The stream to pass to this object. </p></li>
|
||
<li><p><strong>device</strong> – The device on which the stream was created. </p></li>
|
||
<li><p><strong>ownsStream</strong> – Whether this object owns the stream and destroys it in the destructor. </p></li>
|
||
</ul>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10CudaStream10CudaStreamE12cudaStream_t">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10CudaStream10CudaStreamE12cudaStream_t"></span><span id="_CPPv2N12tensorrt_llm7runtime10CudaStream10CudaStreamE12cudaStream_t"></span><span id="tensorrt_llm::runtime::CudaStream::CudaStream__cudaStream_t"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaStream_1a880dbbd2bd43cbf8022969311f8f7d55"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">CudaStream</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">cudaStream_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">stream</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10CudaStream10CudaStreamE12cudaStream_t" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Construct with an existing cuda stream or the default stream by passing nullptr. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10CudaStream9getDeviceEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10CudaStream9getDeviceEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime10CudaStream9getDeviceEv"></span><span id="tensorrt_llm::runtime::CudaStream::getDeviceC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaStream_1a2209ae12d3b5a27740d66bec35e686c9"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">int</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getDevice</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10CudaStream9getDeviceEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Returns the device on which the stream was created. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10CudaStream3getEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10CudaStream3getEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime10CudaStream3getEv"></span><span id="tensorrt_llm::runtime::CudaStream::getC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaStream_1a9ba739eedb71b18bc4a58ff480c82a27"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">cudaStream_t</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">get</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10CudaStream3getEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Returns the stream associated with this object. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10CudaStream11synchronizeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10CudaStream11synchronizeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime10CudaStream11synchronizeEv"></span><span id="tensorrt_llm::runtime::CudaStream::synchronizeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaStream_1a3e3def9cc0e09e9724e9e68232ed2679"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">synchronize</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10CudaStream11synchronizeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Synchronizes the stream. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10CudaStream6recordEN9CudaEvent7pointerE">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10CudaStream6recordEN9CudaEvent7pointerE"></span><span id="_CPPv2NK12tensorrt_llm7runtime10CudaStream6recordEN9CudaEvent7pointerE"></span><span id="tensorrt_llm::runtime::CudaStream::record__CudaEvent::pointerC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaStream_1acc03555968e2361dc08d28f2228a1e31"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">record</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEventE" title="tensorrt_llm::runtime::CudaEvent"><span class="n"><span class="pre">CudaEvent</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEvent7pointerE" title="tensorrt_llm::runtime::CudaEvent::pointer"><span class="n"><span class="pre">pointer</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">event</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10CudaStream6recordEN9CudaEvent7pointerE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Record an event on the stream. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10CudaStream6recordERK9CudaEvent">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10CudaStream6recordERK9CudaEvent"></span><span id="_CPPv2NK12tensorrt_llm7runtime10CudaStream6recordERK9CudaEvent"></span><span id="tensorrt_llm::runtime::CudaStream::record__CudaEventCRC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaStream_1ace6075f1266bdefe0b2033717dd8b14a"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">record</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEventE" title="tensorrt_llm::runtime::CudaEvent"><span class="n"><span class="pre">CudaEvent</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">event</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10CudaStream6recordERK9CudaEvent" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Record an event on the stream. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10CudaStream4waitEN9CudaEvent7pointerE">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10CudaStream4waitEN9CudaEvent7pointerE"></span><span id="_CPPv2NK12tensorrt_llm7runtime10CudaStream4waitEN9CudaEvent7pointerE"></span><span id="tensorrt_llm::runtime::CudaStream::wait__CudaEvent::pointerC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaStream_1a97622ff55b119c5eec1b096115462098"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">wait</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEventE" title="tensorrt_llm::runtime::CudaEvent"><span class="n"><span class="pre">CudaEvent</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEvent7pointerE" title="tensorrt_llm::runtime::CudaEvent::pointer"><span class="n"><span class="pre">pointer</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">event</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10CudaStream4waitEN9CudaEvent7pointerE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Wait for an event. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10CudaStream4waitERK9CudaEvent">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10CudaStream4waitERK9CudaEvent"></span><span id="_CPPv2NK12tensorrt_llm7runtime10CudaStream4waitERK9CudaEvent"></span><span id="tensorrt_llm::runtime::CudaStream::wait__CudaEventCRC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaStream_1a60cfbc942ee8d5f787ae66f891d91766"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">wait</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEventE" title="tensorrt_llm::runtime::CudaEvent"><span class="n"><span class="pre">CudaEvent</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">event</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10CudaStream4waitERK9CudaEvent" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Wait for an event. </p>
|
||
</dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-types">Private Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10CudaStream9StreamPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10CudaStream9StreamPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime10CudaStream9StreamPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaStream_1a42e6eb53f179659c51fbd8981383af64"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">StreamPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">unique_ptr</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">remove_pointer_t</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">cudaStream_t</span></span><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10CudaStream7DeleterE" title="tensorrt_llm::runtime::CudaStream::Deleter"><span class="n"><span class="pre">Deleter</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10CudaStream9StreamPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-members">Private Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10CudaStream7mStreamE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10CudaStream7mStreamE"></span><span id="_CPPv2N12tensorrt_llm7runtime10CudaStream7mStreamE"></span><span id="tensorrt_llm::runtime::CudaStream::mStream__StreamPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaStream_1a21588e234ee724b69f92d1ef26779334"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10CudaStream9StreamPtrE" title="tensorrt_llm::runtime::CudaStream::StreamPtr"><span class="n"><span class="pre">StreamPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mStream</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10CudaStream7mStreamE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10CudaStream7mDeviceE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10CudaStream7mDeviceE"></span><span id="_CPPv2N12tensorrt_llm7runtime10CudaStream7mDeviceE"></span><span id="tensorrt_llm::runtime::CudaStream::mDevice__i"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaStream_1a5dc4aea4b9abfa576f166ce152b675e0"></span><span class="kt"><span class="pre">int</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mDevice</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="o"><span class="pre">-</span></span><span class="m"><span class="pre">1</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10CudaStream7mDeviceE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-friends">Friends</p>
|
||
<dl>
|
||
<dt class="sig sig-object cpp">
|
||
<em class="property"><span class="pre">friend</span> <span class="pre">class</span></em> <span class="pre">CudaStreamBindings</span></dt>
|
||
</dl>
|
||
|
||
</div>
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10CudaStream7DeleterE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10CudaStream7DeleterE"></span><span id="_CPPv2N12tensorrt_llm7runtime10CudaStream7DeleterE"></span><span id="tensorrt_llm::runtime::CudaStream::Deleter"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaStream_1_1Deleter"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Deleter</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10CudaStream7DeleterE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10CudaStream7Deleter7DeleterEb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10CudaStream7Deleter7DeleterEb"></span><span id="_CPPv2N12tensorrt_llm7runtime10CudaStream7Deleter7DeleterEb"></span><span id="tensorrt_llm::runtime::CudaStream::Deleter::Deleter__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaStream_1_1Deleter_1aa03f43a2e696a0433dd1a1c14ce129c2"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Deleter</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">ownsStream</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10CudaStream7Deleter7DeleterEb" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10CudaStream7Deleter7DeleterEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10CudaStream7Deleter7DeleterEv"></span><span id="_CPPv2N12tensorrt_llm7runtime10CudaStream7Deleter7DeleterEv"></span><span id="tensorrt_llm::runtime::CudaStream::Deleter::Deleter"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaStream_1_1Deleter_1adae80ccd16ccabd753f118cdb4111a51"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Deleter</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10CudaStream7Deleter7DeleterEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10CudaStream7DeleterclE12cudaStream_t">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10CudaStream7DeleterclE12cudaStream_t"></span><span id="_CPPv2NK12tensorrt_llm7runtime10CudaStream7DeleterclE12cudaStream_t"></span><span id="tensorrt_llm::runtime::CudaStream::Deleter::call-operator__cudaStream_tCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaStream_1_1Deleter_1a077db80f4a8d30a5f2e0adde1ec6372a"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="k"><span class="pre">operator</span></span><span class="o"><span class="pre">()</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">cudaStream_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">stream</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10CudaStream7DeleterclE12cudaStream_t" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-members">Private Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10CudaStream7Deleter11mOwnsStreamE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10CudaStream7Deleter11mOwnsStreamE"></span><span id="_CPPv2N12tensorrt_llm7runtime10CudaStream7Deleter11mOwnsStreamE"></span><span id="tensorrt_llm::runtime::CudaStream::Deleter::mOwnsStream__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1CudaStream_1_1Deleter_1a71b80a1570697e35d7b5edb51904d9bd"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mOwnsStream</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10CudaStream7Deleter11mOwnsStreamE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="decodinginput-h">
|
||
<h2>decodingInput.h<a class="headerlink" href="#decodinginput-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInputE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInputE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInputE"></span><span id="tensorrt_llm::runtime::DecodingInput"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">DecodingInput</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInputE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="docutils container">
|
||
<em>#include <decodingInput.h></em></div>
|
||
<p>Represents the inputs to the decoder. </p>
|
||
<p>This input type is assumed immutable. It represents whatever the decoder received initially, and can always be referred to as such. </p>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1a051f24825db26577ef03a898c41ee9a0"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorConstPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor14SharedConstPtrE" title="tensorrt_llm::runtime::ITensor::SharedConstPtr"><span class="n"><span class="pre">SharedConstPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1a0e0863b2f0681e5b61953b61b2b072ee"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput9TensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput13DecodingInputE10SizeType3210SizeType3210SizeType3210SizeType3214TensorConstPtr9TensorPtr14TensorConstPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput13DecodingInputE10SizeType3210SizeType3210SizeType3210SizeType3214TensorConstPtr9TensorPtr14TensorConstPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput13DecodingInputE10SizeType3210SizeType3210SizeType3210SizeType3214TensorConstPtr9TensorPtr14TensorConstPtr"></span><span id="tensorrt_llm::runtime::DecodingInput::DecodingInput__SizeType32.SizeType32.SizeType32.SizeType32.TensorConstPtr.TensorPtr.TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1af1901688d26b4be46f628c08d80f447e"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">DecodingInput</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxLength</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxAttentionWindow</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">sinkTokenLength</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">logits</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput9TensorPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">endIds</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchSlots</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput13DecodingInputE10SizeType3210SizeType3210SizeType3210SizeType3214TensorConstPtr9TensorPtr14TensorConstPtr" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput4stepE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput4stepE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput4stepE"></span><span id="tensorrt_llm::runtime::DecodingInput::step__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1a41c49960bea968b4c4e6b7d073c57769"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">step</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput4stepE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>The index of the decoding step we are on. Only used in Python runtime. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput9maxLengthE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput9maxLengthE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput9maxLengthE"></span><span id="tensorrt_llm::runtime::DecodingInput::maxLength__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1abef240110c77063b264d9def9ae87706"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">maxLength</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput9maxLengthE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>The maximum number of tokens to decode. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput18maxAttentionWindowE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput18maxAttentionWindowE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput18maxAttentionWindowE"></span><span id="tensorrt_llm::runtime::DecodingInput::maxAttentionWindow__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1a54ab119d37f9a33cd54c4f9df3db6423"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">maxAttentionWindow</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput18maxAttentionWindowE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>The maximum length of the attention window to consider while decoding. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput15sinkTokenLengthE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput15sinkTokenLengthE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput15sinkTokenLengthE"></span><span id="tensorrt_llm::runtime::DecodingInput::sinkTokenLength__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1a851a4ee559af06eeb0493627d3b8a57f"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">sinkTokenLength</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput15sinkTokenLengthE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>the number of tokens to use as attention sinks, as described there: https://arxiv.org/html/2309.17453v3</p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput9batchSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput9batchSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput9batchSizeE"></span><span id="tensorrt_llm::runtime::DecodingInput::batchSize__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1ab57b39faa8bcf0aa3787a581772e97c3"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">batchSize</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput9batchSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>The number of samples in the batch. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput15maxStopWordsLenE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput15maxStopWordsLenE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput15maxStopWordsLenE"></span><span id="tensorrt_llm::runtime::DecodingInput::maxStopWordsLen__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1a88be1c6c33b42189c86ae0135d042531"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">maxStopWordsLen</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput15maxStopWordsLenE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>The maximum value in the <code class="docutils literal notranslate"><span class="pre">stopWordsLens</span></code> tensor. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput14maxBadWordsLenE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput14maxBadWordsLenE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput14maxBadWordsLenE"></span><span id="tensorrt_llm::runtime::DecodingInput::maxBadWordsLen__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1af0e5d6ebbb1e5dc5fed3ae6c6ac4ca2e"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">maxBadWordsLen</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14maxBadWordsLenE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>The maximum value in the <code class="docutils literal notranslate"><span class="pre">badWordsLens</span></code> tensor. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput6logitsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput6logitsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput6logitsE"></span><span id="tensorrt_llm::runtime::DecodingInput::logits__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1a57cf6d318b6692e9fad94d5cfae221a3"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">logits</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput6logitsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize, beamWidth, vocabSizePadded], on gpu. Logits are are a probability distribution over the vocabulary, the output of the model. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput6endIdsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput6endIdsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput6endIdsE"></span><span id="tensorrt_llm::runtime::DecodingInput::endIds__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1aa493d476a79110129048fe61ba343b0f"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">endIds</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput6endIdsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize * beamWidth], on gpu </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput10batchSlotsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput10batchSlotsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput10batchSlotsE"></span><span id="tensorrt_llm::runtime::DecodingInput::batchSlots__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1a6eb7e6db9122e600018d2ab58a8647b0"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">batchSlots</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput10batchSlotsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize], address map of the linear batch id to to the seq slots, int32_t, pinned </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput13finishReasonsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput13finishReasonsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput13finishReasonsE"></span><span id="tensorrt_llm::runtime::DecodingInput::finishReasons__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1a7bcfc1dcf4652972d5d9b9e2926614c5"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">finishReasons</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput13finishReasonsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize, beamWidth], finished states at current iteration. If true for some request, the decoding step of it is skipped, on gpu </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput19sequenceLimitLengthE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput19sequenceLimitLengthE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput19sequenceLimitLengthE"></span><span id="tensorrt_llm::runtime::DecodingInput::sequenceLimitLength__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1ae6e1f98f774d7800fb5e8c18bf08a74f"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">sequenceLimitLength</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput19sequenceLimitLengthE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize], on gpu. The maximum sequence length for each sequence in the batch. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput13embeddingBiasE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput13embeddingBiasE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput13embeddingBiasE"></span><span id="tensorrt_llm::runtime::DecodingInput::embeddingBias__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1ac7ae35915523b0cae76b9a628d2f8755"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">embeddingBias</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput13embeddingBiasE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize, vocabSizePadded], on gpu </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput7lengthsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput7lengthsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput7lengthsE"></span><span id="tensorrt_llm::runtime::DecodingInput::lengths__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1a6adc71efd0aa8cc7bc3430204b4e71a8"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">lengths</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput7lengthsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize, beamWidth], on gpu </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput13badWordsListsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput13badWordsListsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput13badWordsListsE"></span><span id="tensorrt_llm::runtime::DecodingInput::badWordsLists__std::vector:TensorPtr:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1a88c9c583c32c2f1c3b36f7f426a5b369"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput9TensorPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">badWordsLists</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput13badWordsListsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput12badWordsPtrsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput12badWordsPtrsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput12badWordsPtrsE"></span><span id="tensorrt_llm::runtime::DecodingInput::badWordsPtrs__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1afa6cbf9f8703ccf8bfedd7f24358cdd4"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">badWordsPtrs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput12badWordsPtrsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize][2, badWordsLength], on gpu </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput12badWordsLensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput12badWordsLensE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput12badWordsLensE"></span><span id="tensorrt_llm::runtime::DecodingInput::badWordsLens__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1a1c15b3be4546d48f3e508abaf8f5afce"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">badWordsLens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput12badWordsLensE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize], on gpu </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput14stopWordsListsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput14stopWordsListsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput14stopWordsListsE"></span><span id="tensorrt_llm::runtime::DecodingInput::stopWordsLists__std::vector:TensorPtr:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1a7badda2a04bb863a12c0d22381844c44"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput9TensorPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">stopWordsLists</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14stopWordsListsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput13stopWordsPtrsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput13stopWordsPtrsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput13stopWordsPtrsE"></span><span id="tensorrt_llm::runtime::DecodingInput::stopWordsPtrs__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1a1a912305ebcd3788cc484b51ae6d97d7"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">stopWordsPtrs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput13stopWordsPtrsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize][2, stopWordsLength], pinned </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput13stopWordsLensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput13stopWordsLensE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput13stopWordsLensE"></span><span id="tensorrt_llm::runtime::DecodingInput::stopWordsLens__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1a830d90aeb7e6facbb8195d8cca055ccb"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">stopWordsLens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput13stopWordsLensE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize], pinned </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput17noRepeatNgramSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput17noRepeatNgramSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput17noRepeatNgramSizeE"></span><span id="tensorrt_llm::runtime::DecodingInput::noRepeatNgramSize__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1ae80128aeb288e4aab05278ca2e2512bd"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">noRepeatNgramSize</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput17noRepeatNgramSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize], on gpu </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput16cacheIndirectionE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput16cacheIndirectionE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput16cacheIndirectionE"></span><span id="tensorrt_llm::runtime::DecodingInput::cacheIndirection__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1a6eccbda69f8c35e1eda57e9eb24ca930"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput9TensorPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">cacheIndirection</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput16cacheIndirectionE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize, beamWidth, maxSeqLen] - the k/v cache index for beam search, on gpu </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput12medusaInputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput12medusaInputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput12medusaInputsE"></span><span id="tensorrt_llm::runtime::DecodingInput::medusaInputs__std::optional:MedusaInputs:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1ad5cf2f6414e990319fa0ffaf3a95203a"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput12MedusaInputsE" title="tensorrt_llm::runtime::DecodingInput::MedusaInputs"><span class="n"><span class="pre">MedusaInputs</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">medusaInputs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput12medusaInputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25explicitDraftTokensInputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25explicitDraftTokensInputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25explicitDraftTokensInputsE"></span><span id="tensorrt_llm::runtime::DecodingInput::explicitDraftTokensInputs__std::optional:ExplicitDraftTokensInputs:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1a7bd2778e96e29dafa69f792309aa6046"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputsE" title="tensorrt_llm::runtime::DecodingInput::ExplicitDraftTokensInputs"><span class="n"><span class="pre">ExplicitDraftTokensInputs</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">explicitDraftTokensInputs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25explicitDraftTokensInputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput15lookaheadInputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput15lookaheadInputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput15lookaheadInputsE"></span><span id="tensorrt_llm::runtime::DecodingInput::lookaheadInputs__std::optional:LookaheadInputs:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1ad6524d087f1dfe1873c4f75d4007d9f9"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput15LookaheadInputsE" title="tensorrt_llm::runtime::DecodingInput::LookaheadInputs"><span class="n"><span class="pre">LookaheadInputs</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">lookaheadInputs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput15lookaheadInputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25externalDraftTokensInputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25externalDraftTokensInputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25externalDraftTokensInputsE"></span><span id="tensorrt_llm::runtime::DecodingInput::externalDraftTokensInputs__std::optional:ExternalDraftTokensInputs:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1a2800c0b814e6287f14f1d52098e66282"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputsE" title="tensorrt_llm::runtime::DecodingInput::ExternalDraftTokensInputs"><span class="n"><span class="pre">ExternalDraftTokensInputs</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">externalDraftTokensInputs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25externalDraftTokensInputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput11eagleInputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput11eagleInputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput11eagleInputsE"></span><span id="tensorrt_llm::runtime::DecodingInput::eagleInputs__std::optional:EagleInputs:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1affb7d2c439ced1b7adcf8d00aca41947"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputsE" title="tensorrt_llm::runtime::DecodingInput::EagleInputs"><span class="n"><span class="pre">EagleInputs</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">eagleInputs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput11eagleInputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput11EagleInputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput11EagleInputsE"></span><span id="tensorrt_llm::runtime::DecodingInput::EagleInputs"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DecodingInput_1_1EagleInputs"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">EagleInputs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputs11EagleInputsE14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput11EagleInputs11EagleInputsE14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput11EagleInputs11EagleInputsE14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr"></span><span id="tensorrt_llm::runtime::DecodingInput::EagleInputs::EagleInputs__TensorConstPtr.TensorConstPtr.TensorConstPtr.TensorConstPtr.TensorConstPtr.TensorConstPtr.TensorConstPtr.TensorConstPtr.TensorConstPtr.TensorConstPtr.TensorConstPtr"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DecodingInput_1_1EagleInputs_1a097428c4d295942c0117cd8061ceccf4"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">EagleInputs</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">nextDraftTokens</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">nextDraftLens</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">nextDraftPaths</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">lastDraftTokens</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">lastDraftLens</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">lastDraftPaths</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">acceptedTokens</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">acceptedLens</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">acceptedPathIds</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">chunkedContextNextTokens</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">seqSlots</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputs11EagleInputsE14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr14TensorConstPtr" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputs15nextDraftTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput11EagleInputs15nextDraftTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput11EagleInputs15nextDraftTokensE"></span><span id="tensorrt_llm::runtime::DecodingInput::EagleInputs::nextDraftTokens__TensorConstPtr"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DecodingInput_1_1EagleInputs_1a9d5c2128f6987a534f7857069f4df44d"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">nextDraftTokens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputs15nextDraftTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize, maxDecodingDraftTokens] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputs13nextDraftLensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput11EagleInputs13nextDraftLensE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput11EagleInputs13nextDraftLensE"></span><span id="tensorrt_llm::runtime::DecodingInput::EagleInputs::nextDraftLens__TensorConstPtr"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DecodingInput_1_1EagleInputs_1a965fb843227f57cb06d2ee45791b773d"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">nextDraftLens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputs13nextDraftLensE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputs14nextDraftPathsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput11EagleInputs14nextDraftPathsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput11EagleInputs14nextDraftPathsE"></span><span id="tensorrt_llm::runtime::DecodingInput::EagleInputs::nextDraftPaths__TensorConstPtr"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DecodingInput_1_1EagleInputs_1a7b767bcb8ac75ab7d401a576d6223ab3"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">nextDraftPaths</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputs14nextDraftPathsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize, maxDecodingTokens, maxPathLen] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputs15lastDraftTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput11EagleInputs15lastDraftTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput11EagleInputs15lastDraftTokensE"></span><span id="tensorrt_llm::runtime::DecodingInput::EagleInputs::lastDraftTokens__TensorConstPtr"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DecodingInput_1_1EagleInputs_1a9fa43bad08b57f2650886493c6523631"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">lastDraftTokens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputs15lastDraftTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize, maxNumPaths, maxPathLen] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputs13lastDraftLensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput11EagleInputs13lastDraftLensE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput11EagleInputs13lastDraftLensE"></span><span id="tensorrt_llm::runtime::DecodingInput::EagleInputs::lastDraftLens__TensorConstPtr"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DecodingInput_1_1EagleInputs_1a9250c2234d9562e1a2c727c04897c9c7"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">lastDraftLens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputs13lastDraftLensE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputs14lastDraftPathsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput11EagleInputs14lastDraftPathsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput11EagleInputs14lastDraftPathsE"></span><span id="tensorrt_llm::runtime::DecodingInput::EagleInputs::lastDraftPaths__TensorConstPtr"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DecodingInput_1_1EagleInputs_1a93c600e33b2ccae8b113771bb724869c"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">lastDraftPaths</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputs14lastDraftPathsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize, maxDecodingTokens, maxPathLen] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputs14acceptedTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput11EagleInputs14acceptedTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput11EagleInputs14acceptedTokensE"></span><span id="tensorrt_llm::runtime::DecodingInput::EagleInputs::acceptedTokens__TensorConstPtr"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DecodingInput_1_1EagleInputs_1a84333d7cd56cb5dac547f2c277bbfd9e"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">acceptedTokens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputs14acceptedTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize, maxPathLen] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputs12acceptedLensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput11EagleInputs12acceptedLensE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput11EagleInputs12acceptedLensE"></span><span id="tensorrt_llm::runtime::DecodingInput::EagleInputs::acceptedLens__TensorConstPtr"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DecodingInput_1_1EagleInputs_1aabb443557155e294c75d283b5ca5776f"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">acceptedLens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputs12acceptedLensE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputs15acceptedPathIdsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput11EagleInputs15acceptedPathIdsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput11EagleInputs15acceptedPathIdsE"></span><span id="tensorrt_llm::runtime::DecodingInput::EagleInputs::acceptedPathIds__TensorConstPtr"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DecodingInput_1_1EagleInputs_1a4047afccdec273441c71aff5f259d5bf"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">acceptedPathIds</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputs15acceptedPathIdsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputs24chunkedContextNextTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput11EagleInputs24chunkedContextNextTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput11EagleInputs24chunkedContextNextTokensE"></span><span id="tensorrt_llm::runtime::DecodingInput::EagleInputs::chunkedContextNextTokens__TensorConstPtr"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DecodingInput_1_1EagleInputs_1a7fd68f13a66f6f2e8c814a7700fb9056"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">chunkedContextNextTokens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputs24chunkedContextNextTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputs8seqSlotsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput11EagleInputs8seqSlotsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput11EagleInputs8seqSlotsE"></span><span id="tensorrt_llm::runtime::DecodingInput::EagleInputs::seqSlots__TensorConstPtr"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DecodingInput_1_1EagleInputs_1a2fb8510fab43cee499a8179534fab3d2"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">seqSlots</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput11EagleInputs8seqSlotsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize] </p>
|
||
</dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputsE"></span><span id="tensorrt_llm::runtime::DecodingInput::ExplicitDraftTokensInputs"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1ExplicitDraftTokensInputs"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">ExplicitDraftTokensInputs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs15nextDraftTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs15nextDraftTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs15nextDraftTokensE"></span><span id="tensorrt_llm::runtime::DecodingInput::ExplicitDraftTokensInputs::nextDraftTokens__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1ExplicitDraftTokensInputs_1ab33eb98ffb56f34db936916707a02658"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">nextDraftTokens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs15nextDraftTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize, maxNumPaths, maxPathLen] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs14nextFlatTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs14nextFlatTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs14nextFlatTokensE"></span><span id="tensorrt_llm::runtime::DecodingInput::ExplicitDraftTokensInputs::nextFlatTokens__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1ExplicitDraftTokensInputs_1a7ad74fb97965e08bb1a73cd19a45d14b"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">nextFlatTokens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs14nextFlatTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize * maxDecodingTokens] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs16nextDraftIndicesE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs16nextDraftIndicesE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs16nextDraftIndicesE"></span><span id="tensorrt_llm::runtime::DecodingInput::ExplicitDraftTokensInputs::nextDraftIndices__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1ExplicitDraftTokensInputs_1accc9adb18c4d965102d87fa2b630b277"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">nextDraftIndices</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs16nextDraftIndicesE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize, maxNumPaths, maxPathLen] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs14nextDraftProbsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs14nextDraftProbsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs14nextDraftProbsE"></span><span id="tensorrt_llm::runtime::DecodingInput::ExplicitDraftTokensInputs::nextDraftProbs__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1ExplicitDraftTokensInputs_1a831ac5a03a692eaf6e34cb447e5e8301"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">nextDraftProbs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs14nextDraftProbsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize, maxNumPaths, maxDraftPathLen, vocabSize] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs15lastDraftTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs15lastDraftTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs15lastDraftTokensE"></span><span id="tensorrt_llm::runtime::DecodingInput::ExplicitDraftTokensInputs::lastDraftTokens__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1ExplicitDraftTokensInputs_1a9536a95353e9da425f7d3239765a7ac8"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">lastDraftTokens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs15lastDraftTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize, maxNumPaths, maxPathLen] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs16lastDraftIndicesE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs16lastDraftIndicesE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs16lastDraftIndicesE"></span><span id="tensorrt_llm::runtime::DecodingInput::ExplicitDraftTokensInputs::lastDraftIndices__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1ExplicitDraftTokensInputs_1a9b35756e07900384197581a3b91aeb62"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">lastDraftIndices</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs16lastDraftIndicesE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize, maxNumPaths, maxPathLen] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs5masksE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs5masksE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs5masksE"></span><span id="tensorrt_llm::runtime::DecodingInput::ExplicitDraftTokensInputs::masks__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1ExplicitDraftTokensInputs_1a1a02a579fcf08853b3c115771935e568"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">masks</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs5masksE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize, maxDecodingTokens, maxDecodingTokens], bool </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs17packedPositionIdsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs17packedPositionIdsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs17packedPositionIdsE"></span><span id="tensorrt_llm::runtime::DecodingInput::ExplicitDraftTokensInputs::packedPositionIds__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1ExplicitDraftTokensInputs_1a916530fdfd03ed41753e183c068f2754"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">packedPositionIds</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs17packedPositionIdsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize * maxDecodingTokens] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs15bestPathLengthsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs15bestPathLengthsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs15bestPathLengthsE"></span><span id="tensorrt_llm::runtime::DecodingInput::ExplicitDraftTokensInputs::bestPathLengths__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1ExplicitDraftTokensInputs_1a14ee72c4ff8a12bddb2cc0e2145ef127"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">bestPathLengths</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs15bestPathLengthsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs15bestPathIndicesE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs15bestPathIndicesE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs15bestPathIndicesE"></span><span id="tensorrt_llm::runtime::DecodingInput::ExplicitDraftTokensInputs::bestPathIndices__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1ExplicitDraftTokensInputs_1ad21ef1fb71e9f00a6f67a95086b38deb"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">bestPathIndices</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs15bestPathIndicesE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs21nextGenerationLengthsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs21nextGenerationLengthsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs21nextGenerationLengthsE"></span><span id="tensorrt_llm::runtime::DecodingInput::ExplicitDraftTokensInputs::nextGenerationLengths__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1ExplicitDraftTokensInputs_1a974f61b17e2232c378d2939b08c7507d"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">nextGenerationLengths</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs21nextGenerationLengthsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs19lastPositionIdsBaseE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs19lastPositionIdsBaseE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs19lastPositionIdsBaseE"></span><span id="tensorrt_llm::runtime::DecodingInput::ExplicitDraftTokensInputs::lastPositionIdsBase__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1ExplicitDraftTokensInputs_1a50002e1cd1d62e167745cc694fefb451"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">lastPositionIdsBase</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs19lastPositionIdsBaseE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs21lastGenerationLengthsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs21lastGenerationLengthsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs21lastGenerationLengthsE"></span><span id="tensorrt_llm::runtime::DecodingInput::ExplicitDraftTokensInputs::lastGenerationLengths__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1ExplicitDraftTokensInputs_1ad10825bb06f20d4ce8f67a630c0e04e4"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">lastGenerationLengths</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs21lastGenerationLengthsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs18maxGenLengthDeviceE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs18maxGenLengthDeviceE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs18maxGenLengthDeviceE"></span><span id="tensorrt_llm::runtime::DecodingInput::ExplicitDraftTokensInputs::maxGenLengthDevice__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1ExplicitDraftTokensInputs_1ac1fd1f48b7668c9f9f295c9ad8ccfad0"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">maxGenLengthDevice</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs18maxGenLengthDeviceE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[1] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs8seqSlotsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs8seqSlotsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs8seqSlotsE"></span><span id="tensorrt_llm::runtime::DecodingInput::ExplicitDraftTokensInputs::seqSlots__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1ExplicitDraftTokensInputs_1abfd632f47aff831cd43f55eb40b1ea82"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">seqSlots</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExplicitDraftTokensInputs8seqSlotsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize] </p>
|
||
</dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputsE"></span><span id="tensorrt_llm::runtime::DecodingInput::ExternalDraftTokensInputs"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1ExternalDraftTokensInputs"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">ExternalDraftTokensInputs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs11draftLogitsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs11draftLogitsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs11draftLogitsE"></span><span id="tensorrt_llm::runtime::DecodingInput::ExternalDraftTokensInputs::draftLogits__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1ExternalDraftTokensInputs_1addf6ccc0d11838a16a36d4d574f2149b"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput9TensorPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">draftLogits</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs11draftLogitsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs10draftProbsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs10draftProbsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs10draftProbsE"></span><span id="tensorrt_llm::runtime::DecodingInput::ExternalDraftTokensInputs::draftProbs__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1ExternalDraftTokensInputs_1ab39186249926f9f16fc399f1f47db321"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput9TensorPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">draftProbs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs10draftProbsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs11targetProbsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs11targetProbsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs11targetProbsE"></span><span id="tensorrt_llm::runtime::DecodingInput::ExternalDraftTokensInputs::targetProbs__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1ExternalDraftTokensInputs_1a1257278db85997fab8bc0c6d604e95c1"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput9TensorPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">targetProbs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs11targetProbsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs14numDraftTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs14numDraftTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs14numDraftTokensE"></span><span id="tensorrt_llm::runtime::DecodingInput::ExternalDraftTokensInputs::numDraftTokens__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1ExternalDraftTokensInputs_1a0d5d346e1f80a3bac75fcab5b7c58369"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput9TensorPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">numDraftTokens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs14numDraftTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs13draftTokenIdsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs13draftTokenIdsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs13draftTokenIdsE"></span><span id="tensorrt_llm::runtime::DecodingInput::ExternalDraftTokensInputs::draftTokenIds__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1ExternalDraftTokensInputs_1a8ff083621c48c2ef0a74847f17925c6a"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput9TensorPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">draftTokenIds</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs13draftTokenIdsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs14useDraftLogitsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs14useDraftLogitsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs14useDraftLogitsE"></span><span id="tensorrt_llm::runtime::DecodingInput::ExternalDraftTokensInputs::useDraftLogits__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1ExternalDraftTokensInputs_1ae15a164a475300b1e268048fa080c00e"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput9TensorPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">useDraftLogits</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs14useDraftLogitsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs18useDraftLogitsHostE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs18useDraftLogitsHostE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs18useDraftLogitsHostE"></span><span id="tensorrt_llm::runtime::DecodingInput::ExternalDraftTokensInputs::useDraftLogitsHost__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1ExternalDraftTokensInputs_1a5223b07547875357d19c7e9bf2bf0c9b"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput9TensorPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">useDraftLogitsHost</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs18useDraftLogitsHostE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs4stepE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs4stepE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs4stepE"></span><span id="tensorrt_llm::runtime::DecodingInput::ExternalDraftTokensInputs::step__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1ExternalDraftTokensInputs_1a7d222ffc1f7db7497ce4a809bcf3779a"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">step</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs4stepE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs17constantThresholdE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs17constantThresholdE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs17constantThresholdE"></span><span id="tensorrt_llm::runtime::DecodingInput::ExternalDraftTokensInputs::constantThreshold__float"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1ExternalDraftTokensInputs_1ad6952502823a4452f686bc2c5a574f5b"></span><span class="kt"><span class="pre">float</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">constantThreshold</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs17constantThresholdE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs28useRandomAcceptanceThresholdE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs28useRandomAcceptanceThresholdE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs28useRandomAcceptanceThresholdE"></span><span id="tensorrt_llm::runtime::DecodingInput::ExternalDraftTokensInputs::useRandomAcceptanceThreshold__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1ExternalDraftTokensInputs_1ace748e3667f3462d8edb615c808a78d7"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">useRandomAcceptanceThreshold</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput25ExternalDraftTokensInputs28useRandomAcceptanceThresholdE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput15LookaheadInputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput15LookaheadInputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput15LookaheadInputsE"></span><span id="tensorrt_llm::runtime::DecodingInput::LookaheadInputs"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DecodingInput_1_1LookaheadInputs"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">LookaheadInputs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput15LookaheadInputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput15LookaheadInputs13tokensPerStepE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput15LookaheadInputs13tokensPerStepE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput15LookaheadInputs13tokensPerStepE"></span><span id="tensorrt_llm::runtime::DecodingInput::LookaheadInputs::tokensPerStep__TensorPtr"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DecodingInput_1_1LookaheadInputs_1a646f2bcd543ac02e92b3c8cc40e2c920"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput9TensorPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tokensPerStep</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput15LookaheadInputs13tokensPerStepE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput12MedusaInputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput12MedusaInputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput12MedusaInputsE"></span><span id="tensorrt_llm::runtime::DecodingInput::MedusaInputs"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1MedusaInputs"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">MedusaInputs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput12MedusaInputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput12MedusaInputs11medusaPathsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput12MedusaInputs11medusaPathsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput12MedusaInputs11medusaPathsE"></span><span id="tensorrt_llm::runtime::DecodingInput::MedusaInputs::medusaPaths__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1MedusaInputs_1a5307b7803b035e1ce0814dc8523a8f60"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">medusaPaths</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput12MedusaInputs11medusaPathsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize, maxTokensPerStep, maxMedusaHeads + 1], on gpu </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput12MedusaInputs13medusaTreeIdsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput12MedusaInputs13medusaTreeIdsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput12MedusaInputs13medusaTreeIdsE"></span><span id="tensorrt_llm::runtime::DecodingInput::MedusaInputs::medusaTreeIds__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1MedusaInputs_1ad5ce6d5babb4b1a5c60c1203e9594f87"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">medusaTreeIds</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput12MedusaInputs13medusaTreeIdsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize, maxTokensPerStep], on gpu </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput12MedusaInputs12medusaLogitsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput12MedusaInputs12medusaLogitsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput12MedusaInputs12medusaLogitsE"></span><span id="tensorrt_llm::runtime::DecodingInput::MedusaInputs::medusaLogits__std::vector:std::vector:TensorPtr::"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1MedusaInputs_1a61552ddc3f54658c465929a297912036"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput9TensorPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">medusaLogits</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput12MedusaInputs12medusaLogitsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize][maxAcceptedDraftTokensPerStep][maxDraftTokens + 1, vocabSizePadded], on gpu </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput12MedusaInputs22medusaCurTokensPerStepE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput12MedusaInputs22medusaCurTokensPerStepE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput12MedusaInputs22medusaCurTokensPerStepE"></span><span id="tensorrt_llm::runtime::DecodingInput::MedusaInputs::medusaCurTokensPerStep__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1MedusaInputs_1a1a8c2fcc38cb3c8c3c62ba80bc0cf60f"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput9TensorPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">medusaCurTokensPerStep</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput12MedusaInputs22medusaCurTokensPerStepE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize], on gpu </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput12MedusaInputs25medusaTargetTokensPerStepE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput12MedusaInputs25medusaTargetTokensPerStepE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput12MedusaInputs25medusaTargetTokensPerStepE"></span><span id="tensorrt_llm::runtime::DecodingInput::MedusaInputs::medusaTargetTokensPerStep__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1_1MedusaInputs_1a65af55c74760e3aef0f03177f604e849"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput14TensorConstPtrE" title="tensorrt_llm::runtime::DecodingInput::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">medusaTargetTokensPerStep</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput12MedusaInputs25medusaTargetTokensPerStepE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize], on gpu </p>
|
||
</dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="decodingoutput-h">
|
||
<h2>decodingOutput.h<a class="headerlink" href="#decodingoutput-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm13batch_managerE">
|
||
<span id="_CPPv3N12tensorrt_llm13batch_managerE"></span><span id="_CPPv2N12tensorrt_llm13batch_managerE"></span><span id="tensorrt_llm::batch_manager"></span><span class="target" id="namespacetensorrt__llm_1_1batch__manager"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">batch_manager</span></span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutputE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutputE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutputE"></span><span id="tensorrt_llm::runtime::DecodingOutput"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">DecodingOutput</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutputE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1aeb215aae60278c44bcdbd17ae5f7c8e1"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput14DecodingOutputE9TensorPtr9TensorPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput14DecodingOutputE9TensorPtr9TensorPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput14DecodingOutputE9TensorPtr9TensorPtr"></span><span id="tensorrt_llm::runtime::DecodingOutput::DecodingOutput__TensorPtr.TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1af68b04c734d41e77cea6854abd7bdc2c"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">DecodingOutput</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">ids</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">gatheredIds</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14DecodingOutputE9TensorPtr9TensorPtr" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput3idsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput3idsE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput3idsE"></span><span id="tensorrt_llm::runtime::DecodingOutput::ids__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1a9937790aaf8c5cbb5230236afc7656be"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">ids</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput3idsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput11gatheredIdsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput11gatheredIdsE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput11gatheredIdsE"></span><span id="tensorrt_llm::runtime::DecodingOutput::gatheredIds__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1a7694f98fca6aed5f97e731217d97ff50"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">gatheredIds</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput11gatheredIdsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput14newTokensStepsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput14newTokensStepsE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput14newTokensStepsE"></span><span id="tensorrt_llm::runtime::DecodingOutput::newTokensSteps__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1a6e42e3b0fd2b24ba8a04f78f837b207a"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">newTokensSteps</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14newTokensStepsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput9newTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput9newTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput9newTokensE"></span><span id="tensorrt_llm::runtime::DecodingOutput::newTokens__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1a1c92d23772ed90b78f9ac8c86f46d9fd"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">newTokens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9newTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput12newTokensVecE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput12newTokensVecE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput12newTokensVecE"></span><span id="tensorrt_llm::runtime::DecodingOutput::newTokensVec__std::vector:TensorPtr:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1a0df2485f70ab5959e706840f7635e4c3"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">newTokensVec</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput12newTokensVecE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput13finishReasonsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput13finishReasonsE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput13finishReasonsE"></span><span id="tensorrt_llm::runtime::DecodingOutput::finishReasons__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1a1b520adda489db15ba1b8fde1baa195f"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">finishReasons</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput13finishReasonsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput11finishedSumE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput11finishedSumE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput11finishedSumE"></span><span id="tensorrt_llm::runtime::DecodingOutput::finishedSum__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1a2bb2c2cc930598e59363a8d61e2f0fc3"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">finishedSum</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput11finishedSumE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput8logProbsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput8logProbsE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput8logProbsE"></span><span id="tensorrt_llm::runtime::DecodingOutput::logProbs__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1a4abf75dc398349212b15b93f843fc03c"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">logProbs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput8logProbsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput11cumLogProbsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput11cumLogProbsE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput11cumLogProbsE"></span><span id="tensorrt_llm::runtime::DecodingOutput::cumLogProbs__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1a428f7d981f0617bda37e6f64bb4f0007"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">cumLogProbs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput11cumLogProbsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput9parentIdsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput9parentIdsE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput9parentIdsE"></span><span id="tensorrt_llm::runtime::DecodingOutput::parentIds__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1a9f518ec59bd0df527432074008d974a6"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">parentIds</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9parentIdsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput7lengthsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput7lengthsE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput7lengthsE"></span><span id="tensorrt_llm::runtime::DecodingOutput::lengths__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1a7425f953e704cffb2b917d475fc90c12"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">lengths</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput7lengthsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput16cacheIndirectionE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput16cacheIndirectionE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput16cacheIndirectionE"></span><span id="tensorrt_llm::runtime::DecodingOutput::cacheIndirection__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1a6ee6dfc4bf8f6f7ae17e0fc8d11ea56d"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">cacheIndirection</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput16cacheIndirectionE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput13logProbsTiledE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput13logProbsTiledE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput13logProbsTiledE"></span><span id="tensorrt_llm::runtime::DecodingOutput::logProbsTiled__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1ac0ca8c7f11b53aa97c4caa6282853a65"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">logProbsTiled</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput13logProbsTiledE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput14beamHypothesesE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput14beamHypothesesE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput14beamHypothesesE"></span><span id="tensorrt_llm::runtime::DecodingOutput::beamHypotheses__BeamHypotheses"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1a1e1c48231792a45618deff974b48b4ee"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypothesesE" title="tensorrt_llm::runtime::DecodingOutput::BeamHypotheses"><span class="n"><span class="pre">BeamHypotheses</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">beamHypotheses</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14beamHypothesesE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput26speculativeDecodingOutputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput26speculativeDecodingOutputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput26speculativeDecodingOutputsE"></span><span id="tensorrt_llm::runtime::DecodingOutput::speculativeDecodingOutputs__std::optional:SpeculativeDecodingOutputs:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1a14f7bfd40b8c07bacab271148aabc2d7"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputsE" title="tensorrt_llm::runtime::DecodingOutput::SpeculativeDecodingOutputs"><span class="n"><span class="pre">SpeculativeDecodingOutputs</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">speculativeDecodingOutputs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput26speculativeDecodingOutputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput26explicitDraftTokensBuffersE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput26explicitDraftTokensBuffersE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput26explicitDraftTokensBuffersE"></span><span id="tensorrt_llm::runtime::DecodingOutput::explicitDraftTokensBuffers__std::optional:ExplicitDraftTokensBuffers::Inputs:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1a0379b1ee160c10fcdaeef426c8799a1b"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffersE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers"><span class="n"><span class="pre">ExplicitDraftTokensBuffers</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6InputsE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::Inputs"><span class="n"><span class="pre">Inputs</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">explicitDraftTokensBuffers</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput26explicitDraftTokensBuffersE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput16lookaheadOutputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput16lookaheadOutputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput16lookaheadOutputsE"></span><span id="tensorrt_llm::runtime::DecodingOutput::lookaheadOutputs__std::optional:LookaheadDecodingBuffers:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1a304d012c64f4a46e8fda750824e12063"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffersE" title="tensorrt_llm::runtime::LookaheadDecodingBuffers"><span class="n"><span class="pre">LookaheadDecodingBuffers</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">lookaheadOutputs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput16lookaheadOutputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput12eagleBuffersE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput12eagleBuffersE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput12eagleBuffersE"></span><span id="tensorrt_llm::runtime::DecodingOutput::eagleBuffers__std::optional:EagleBuffers::Inputs:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1a1123acfc58f9842de75bdf8a29316b58"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffersE" title="tensorrt_llm::runtime::EagleBuffers"><span class="n"><span class="pre">EagleBuffers</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6InputsE" title="tensorrt_llm::runtime::EagleBuffers::Inputs"><span class="n"><span class="pre">Inputs</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">eagleBuffers</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput12eagleBuffersE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput17kNegativeInfinityE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput17kNegativeInfinityE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput17kNegativeInfinityE"></span><span id="tensorrt_llm::runtime::DecodingOutput::kNegativeInfinity__float"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1a1c6759c548ba677d9ea18fb8c895a6aa"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">float</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kNegativeInfinity</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="o"><span class="pre">-</span></span><span class="m"><span class="pre">1e20f</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput17kNegativeInfinityE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypothesesE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput14BeamHypothesesE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput14BeamHypothesesE"></span><span id="tensorrt_llm::runtime::DecodingOutput::BeamHypotheses"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1BeamHypotheses"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">BeamHypotheses</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypothesesE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses5emptyER13BufferManager">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses5emptyER13BufferManager"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses5emptyER13BufferManager"></span><span id="tensorrt_llm::runtime::DecodingOutput::BeamHypotheses::empty__BufferManagerR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1BeamHypotheses_1a00c82796e448818892e94f8a064a74cd"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">empty</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManagerE" title="tensorrt_llm::runtime::BufferManager"><span class="n"><span class="pre">BufferManager</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">manager</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses5emptyER13BufferManager" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses7reshapeE10SizeType3210SizeType3210SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses7reshapeE10SizeType3210SizeType3210SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses7reshapeE10SizeType3210SizeType3210SizeType32"></span><span id="tensorrt_llm::runtime::DecodingOutput::BeamHypotheses::reshape__SizeType32.SizeType32.SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1BeamHypotheses_1a481322408d5bcdd80a03cb104bca32be"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">reshape</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">beamWidth</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxSequenceLength</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses7reshapeE10SizeType3210SizeType3210SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses7releaseEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses7releaseEv"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses7releaseEv"></span><span id="tensorrt_llm::runtime::DecodingOutput::BeamHypotheses::release"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1BeamHypotheses_1a6c10ab322d436f29ce99f0f0cb4e1e1a"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">release</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses7releaseEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses4initER13BufferManager11TokenIdType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses4initER13BufferManager11TokenIdType"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses4initER13BufferManager11TokenIdType"></span><span id="tensorrt_llm::runtime::DecodingOutput::BeamHypotheses::init__BufferManagerR.TokenIdType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1BeamHypotheses_1ab42b472a1d818532cd173ba69ad6097e"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">init</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManagerE" title="tensorrt_llm::runtime::BufferManager"><span class="n"><span class="pre">BufferManager</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">manager</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11TokenIdTypeE" title="tensorrt_llm::runtime::TokenIdType"><span class="n"><span class="pre">TokenIdType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">endId</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses4initER13BufferManager11TokenIdType" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses5sliceE10SizeType3210SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses5sliceE10SizeType3210SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses5sliceE10SizeType3210SizeType32"></span><span id="tensorrt_llm::runtime::DecodingOutput::BeamHypotheses::slice__SizeType32.SizeType32C"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1BeamHypotheses_1a50635b6fe49fb91ae5bcfe7f0c49c96e"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypothesesE" title="tensorrt_llm::runtime::DecodingOutput::BeamHypotheses"><span class="n"><span class="pre">BeamHypotheses</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">slice</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchIndex</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses5sliceE10SizeType3210SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses12outputIdsCBAE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses12outputIdsCBAE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses12outputIdsCBAE"></span><span id="tensorrt_llm::runtime::DecodingOutput::BeamHypotheses::outputIdsCBA__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1BeamHypotheses_1a0961829db8fc22b9d37455958b6af513"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">outputIdsCBA</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses12outputIdsCBAE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses11logProbsCBAE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses11logProbsCBAE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses11logProbsCBAE"></span><span id="tensorrt_llm::runtime::DecodingOutput::BeamHypotheses::logProbsCBA__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1BeamHypotheses_1ab4c51ee4b3b57de60a07e2e4e0ef1b69"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">logProbsCBA</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses11logProbsCBAE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses18sequenceLengthsCBAE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses18sequenceLengthsCBAE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses18sequenceLengthsCBAE"></span><span id="tensorrt_llm::runtime::DecodingOutput::BeamHypotheses::sequenceLengthsCBA__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1BeamHypotheses_1afd44fb8972884de1ca81f62ff7a55189"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">sequenceLengthsCBA</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses18sequenceLengthsCBAE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses14cumLogProbsCBAE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses14cumLogProbsCBAE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses14cumLogProbsCBAE"></span><span id="tensorrt_llm::runtime::DecodingOutput::BeamHypotheses::cumLogProbsCBA__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1BeamHypotheses_1a7f0329d9772be169a083bd57fcfe1691"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">cumLogProbsCBA</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses14cumLogProbsCBAE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses15normedScoresCBAE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses15normedScoresCBAE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses15normedScoresCBAE"></span><span id="tensorrt_llm::runtime::DecodingOutput::BeamHypotheses::normedScoresCBA__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1BeamHypotheses_1a66b2659d6b50d9e9ae6fe0e8c10c9d08"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">normedScoresCBA</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses15normedScoresCBAE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses11numBeamsCBAE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses11numBeamsCBAE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses11numBeamsCBAE"></span><span id="tensorrt_llm::runtime::DecodingOutput::BeamHypotheses::numBeamsCBA__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1BeamHypotheses_1a2369a7d4bf929356c3441fb4d9ccdc64"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">numBeamsCBA</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses11numBeamsCBAE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses18minNormedScoresCBAE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses18minNormedScoresCBAE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses18minNormedScoresCBAE"></span><span id="tensorrt_llm::runtime::DecodingOutput::BeamHypotheses::minNormedScoresCBA__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1BeamHypotheses_1a4db33cb17d3de5a6635bf03132af2633"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">minNormedScoresCBA</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses18minNormedScoresCBAE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses10batchDonesE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses10batchDonesE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses10batchDonesE"></span><span id="tensorrt_llm::runtime::DecodingOutput::BeamHypotheses::batchDones__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1BeamHypotheses_1a7d6198e1a65cfea9755483162c8e139e"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">batchDones</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses10batchDonesE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputsE"></span><span id="tensorrt_llm::runtime::DecodingOutput::SpeculativeDecodingOutputs"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1SpeculativeDecodingOutputs"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">SpeculativeDecodingOutputs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputs15nextDraftTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputs15nextDraftTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputs15nextDraftTokensE"></span><span id="tensorrt_llm::runtime::DecodingOutput::SpeculativeDecodingOutputs::nextDraftTokens__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1SpeculativeDecodingOutputs_1a1f368e032acf0d91ee64b338ccdcf32c"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">nextDraftTokens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputs15nextDraftTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputs18nextDraftTokensLenE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputs18nextDraftTokensLenE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputs18nextDraftTokensLenE"></span><span id="tensorrt_llm::runtime::DecodingOutput::SpeculativeDecodingOutputs::nextDraftTokensLen__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1SpeculativeDecodingOutputs_1a69501f3834d3b2251b5fb8d01ec489b6"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">nextDraftTokensLen</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputs18nextDraftTokensLenE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputs18prevDraftTokensLenE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputs18prevDraftTokensLenE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputs18prevDraftTokensLenE"></span><span id="tensorrt_llm::runtime::DecodingOutput::SpeculativeDecodingOutputs::prevDraftTokensLen__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1SpeculativeDecodingOutputs_1a52297748054f430339c310d2ac330bde"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">prevDraftTokensLen</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputs18prevDraftTokensLenE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputs17acceptedTokensLenE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputs17acceptedTokensLenE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputs17acceptedTokensLenE"></span><span id="tensorrt_llm::runtime::DecodingOutput::SpeculativeDecodingOutputs::acceptedTokensLen__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1SpeculativeDecodingOutputs_1ad1b7961868d99497e4234ad4b8f52af8"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">acceptedTokensLen</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputs17acceptedTokensLenE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputs21acceptedLengthsCumSumE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputs21acceptedLengthsCumSumE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputs21acceptedLengthsCumSumE"></span><span id="tensorrt_llm::runtime::DecodingOutput::SpeculativeDecodingOutputs::acceptedLengthsCumSum__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1SpeculativeDecodingOutputs_1ad830ac27b0d38fc08ae08b9a9f408058"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">acceptedLengthsCumSum</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputs21acceptedLengthsCumSumE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputs12pathsOffsetsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputs12pathsOffsetsE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputs12pathsOffsetsE"></span><span id="tensorrt_llm::runtime::DecodingOutput::SpeculativeDecodingOutputs::pathsOffsets__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1SpeculativeDecodingOutputs_1a03210ddbe99a5780c2301d109536294b"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">pathsOffsets</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput26SpeculativeDecodingOutputs12pathsOffsetsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="generationinput-h">
|
||
<h2>generationInput.h<a class="headerlink" href="#generationinput-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GenerationInputE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GenerationInputE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GenerationInputE"></span><span id="tensorrt_llm::runtime::GenerationInput"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenerationInput"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">GenerationInput</span></span></span><span class="w"> </span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="k"><span class="pre">public</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4I00EN12tensorrt_llm7runtime22GenericGenerationInputE" title="tensorrt_llm::runtime::GenericGenerationInput"><span class="n"><span class="pre">GenericGenerationInput</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><span class="p"><span class="pre">,</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18PromptTuningParamsE" title="tensorrt_llm::runtime::PromptTuningParams"><span class="n"><span class="pre">PromptTuningParams</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15GenerationInputE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GenerationInput4BaseE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GenerationInput4BaseE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GenerationInput4BaseE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenerationInput_1ac4231ce739fe28958d51d834024d92cd"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Base</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4I00EN12tensorrt_llm7runtime22GenericGenerationInputE" title="tensorrt_llm::runtime::GenericGenerationInput"><span class="n"><span class="pre">GenericGenerationInput</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><span class="p"><span class="pre">,</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18PromptTuningParamsE" title="tensorrt_llm::runtime::PromptTuningParams"><span class="n"><span class="pre">PromptTuningParams</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15GenerationInput4BaseE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GenerationInput9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GenerationInput9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GenerationInput9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenerationInput_1ad525a655e0d2658c3f99dabb572acf1a"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15GenerationInput4BaseE" title="tensorrt_llm::runtime::GenerationInput::Base"><span class="n"><span class="pre">Base</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">TensorPtr</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15GenerationInput9TensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GenerationInput15GenerationInputEK10SizeType32K10SizeType329TensorPtr9TensorPtrb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GenerationInput15GenerationInputEK10SizeType32K10SizeType329TensorPtr9TensorPtrb"></span><span id="_CPPv2N12tensorrt_llm7runtime15GenerationInput15GenerationInputEK10SizeType32K10SizeType329TensorPtr9TensorPtrb"></span><span id="tensorrt_llm::runtime::GenerationInput::GenerationInput__SizeType32C.SizeType32C.TensorPtr.TensorPtr.b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenerationInput_1a60002f387300ebc1de3bd0b25796353b"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">GenerationInput</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">endId</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">padId</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15GenerationInput9TensorPtrE" title="tensorrt_llm::runtime::GenerationInput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">ids</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15GenerationInput9TensorPtrE" title="tensorrt_llm::runtime::GenerationInput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">lengths</span></span>, <span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">packed</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">false</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15GenerationInput15GenerationInputEK10SizeType32K10SizeType329TensorPtr9TensorPtrb" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I00EN12tensorrt_llm7runtime22GenericGenerationInputE">
|
||
<span id="_CPPv3I00EN12tensorrt_llm7runtime22GenericGenerationInputE"></span><span id="_CPPv2I00EN12tensorrt_llm7runtime22GenericGenerationInputE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TTensor</span></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">PromptTuningParams</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericGenerationInput"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">GenericGenerationInput</span></span></span><a class="headerlink" href="#_CPPv4I00EN12tensorrt_llm7runtime22GenericGenerationInputE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="docutils container">
|
||
<em>#include <generationInput.h></em></div>
|
||
<p><em></em></p>
|
||
<p><ul class="simple">
|
||
<li><p><code class="docutils literal notranslate"><span class="pre">endId</span></code>, is the token ID that marks the end of the input sequence (aka <code class="docutils literal notranslate"><span class="pre">EOS</span></code> or end-of-sequence). It’s <code class="docutils literal notranslate"><span class="pre">50,256</span></code> for the GPT2 model which has a vocabulary of <code class="docutils literal notranslate"><span class="pre">50,257</span></code> tokens, for example,</p></li>
|
||
<li><p><code class="docutils literal notranslate"><span class="pre">padId</span></code>, is the token ID that is used for padding (i.e. fills in the slots that are at an index greater-or-equal to the input length for padded sequences). It can be set to the same value as <code class="docutils literal notranslate"><span class="pre">endId</span></code>,</p></li>
|
||
<li><p><code class="docutils literal notranslate"><span class="pre">ids</span></code>, is the tensor of input IDs. That tensor must be allocated on the GPU. When the input tensor is padded, the shape of <code class="docutils literal notranslate"><span class="pre">ids</span></code> is <code class="docutils literal notranslate"><span class="pre">[batchSize,</span> <span class="pre">maxInputLength]</span></code>, where <code class="docutils literal notranslate"><span class="pre">batchSize</span></code> and <code class="docutils literal notranslate"><span class="pre">maxInputLength</span></code> must respect the maximum sizes in <code class="docutils literal notranslate"><span class="pre">sessionConfig</span></code> passed to the <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1GptSession"><span class="std std-ref"><span class="pre">GptSession</span></span></a></code> constructor. When the input is packed, the shape of <code class="docutils literal notranslate"><span class="pre">ids</span></code> is <code class="docutils literal notranslate"><span class="pre">[numTokens]</span></code>, where <code class="docutils literal notranslate"><span class="pre">numTokens</span></code> is the sum of the lengths of the different sequences in the batch,</p></li>
|
||
<li><p><code class="docutils literal notranslate"><span class="pre">lengths</span></code>, is the tensor of input sequence lengths. That tensor must be allocated on the GPU and contain <code class="docutils literal notranslate"><span class="pre">batchSize</span></code> values,</p></li>
|
||
<li><p><code class="docutils literal notranslate"><span class="pre">packed</span></code>, indicates if the <code class="docutils literal notranslate"><span class="pre">ids</span></code> tensor is packed or padded. In this release, that flag must match the value passed to the constructor through the instance of the <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1ModelConfig"><span class="std std-ref"><span class="pre">ModelConfig</span></span></a></code> class. In a future release, the session may be made more flexible and automatically pad or pack the input,</p></li>
|
||
</ul>
|
||
</p>
|
||
<p><em></em></p>
|
||
<p><ul class="simple">
|
||
<li><p><code class="docutils literal notranslate"><span class="pre">embeddingBiasOpt</span></code>, is a tensor of floating-point values on the GPU that contains the bias to add to the logits during sampling (after the projection from hidden states to logits as the last step of the model). This tensor must have <code class="docutils literal notranslate"><span class="pre">vocabSize</span></code> elements (as defined in the <code class="docutils literal notranslate"><span class="pre">modelConfig</span></code> argument passed to the constructor),</p></li>
|
||
<li><p><code class="docutils literal notranslate"><span class="pre">badWordsList</span></code>, is a tensor of integers on the GPU that encodes the list of words that have to be banned from generated sequences. Its shape is <code class="docutils literal notranslate"><span class="pre">[2,</span> <span class="pre">badWordsLength]</span></code>, as explained below, or <code class="docutils literal notranslate"><span class="pre">[batchSize,</span> <span class="pre">2,</span> <span class="pre">badWordsLength]</span></code> when there is a different list for each sequence in the batch,</p></li>
|
||
<li><p><code class="docutils literal notranslate"><span class="pre">stopWordsList</span></code>, is a tensor of integers on the GPU that encodes the list of words that trigger the end of the generation for a sequence. Its shape is <code class="docutils literal notranslate"><span class="pre">[2,</span> <span class="pre">stopWordsLength]</span></code>, as explained below, or <code class="docutils literal notranslate"><span class="pre">[batchSize,</span> <span class="pre">2,</span> <span class="pre">stopWordsLength]</span></code> when there is a different list for each sequence in the batch,</p></li>
|
||
<li><p><code class="docutils literal notranslate"><span class="pre">maxNewTokens</span></code>, is the maximum number of tokens to generate.</p></li>
|
||
</ul>
|
||
</p>
|
||
<p>The <code class="docutils literal notranslate"><span class="pre">badWordsList</span></code> and <code class="docutils literal notranslate"><span class="pre">stopWordsList</span></code> tensors have the same shape <code class="docutils literal notranslate"><span class="pre">[2,</span> <span class="pre">length]</span></code>. Let’s consider an example with three words to describe the representation of those lists. The first word contains tokens <code class="docutils literal notranslate"><span class="pre">[5,</span> <span class="pre">7,</span> <span class="pre">3]</span></code>, the second one contains <code class="docutils literal notranslate"><span class="pre">[9,</span> <span class="pre">2]</span></code> and the third one is composed of tokens <code class="docutils literal notranslate"><span class="pre">[6,</span> <span class="pre">2,</span> <span class="pre">4,</span> <span class="pre">1]</span></code>. In total, there are 9 tokens. That’s the length. The shape of the tensor is <code class="docutils literal notranslate"><span class="pre">[2,</span> <span class="pre">9]</span></code>. The first row of the tensor must contain the 9 token IDs and the second row must store the <a class="reference external" href="https://en.wikipedia.org/wiki/Prefix_sum">inclusive prefix-sum</a> of the word lengths as shown on the following diagram:</p>
|
||
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span> <span class="mi">0</span> <span class="mi">3</span> <span class="mi">5</span> <span class="mi">9</span>
|
||
<span class="o">|</span> <span class="o">|</span> <span class="o">|</span> <span class="o">|</span>
|
||
<span class="n">V</span> <span class="n">V</span> <span class="n">V</span> <span class="n">V</span>
|
||
<span class="p">[</span> <span class="mi">5</span><span class="p">,</span> <span class="mi">7</span><span class="p">,</span> <span class="mi">3</span><span class="p">,</span> <span class="mi">9</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="mi">6</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="mi">4</span><span class="p">,</span> <span class="mi">1</span><span class="p">]</span>
|
||
<span class="p">[</span> <span class="mi">3</span><span class="p">,</span> <span class="mi">5</span><span class="p">,</span> <span class="mi">9</span><span class="p">,</span> <span class="o">-</span><span class="mi">1</span><span class="p">,</span> <span class="o">-</span><span class="mi">1</span><span class="p">,</span> <span class="o">-</span><span class="mi">1</span><span class="p">,</span> <span class="o">-</span><span class="mi">1</span><span class="p">,</span> <span class="o">-</span><span class="mi">1</span><span class="p">,</span> <span class="o">-</span><span class="mi">1</span><span class="p">]</span>
|
||
</pre></div>
|
||
</div>
|
||
<p>In case all the words are made of a single token, the inner-most dimension of the tensor must be increased by 1 (i.e. the length for 4 words, each made of a single token, must be 5 instead of 4 &#8212; the shape is <code class="docutils literal notranslate"><span class="pre">[2,</span> <span class="pre">5]</span></code>). </p>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime22GenericGenerationInput9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime22GenericGenerationInput9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericGenerationInput_1a9697fefb8d30d140fb950c041cd3c86b"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4I00EN12tensorrt_llm7runtime22GenericGenerationInputE" title="tensorrt_llm::runtime::GenericGenerationInput::TTensor"><span class="n"><span class="pre">TTensor</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput9TensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput22GenericGenerationInputEK10SizeType32K10SizeType329TensorPtr9TensorPtrb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime22GenericGenerationInput22GenericGenerationInputEK10SizeType32K10SizeType329TensorPtr9TensorPtrb"></span><span id="_CPPv2N12tensorrt_llm7runtime22GenericGenerationInput22GenericGenerationInputEK10SizeType32K10SizeType329TensorPtr9TensorPtrb"></span><span id="tensorrt_llm::runtime::GenericGenerationInput::GenericGenerationInput__SizeType32C.SizeType32C.TensorPtr.TensorPtr.b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericGenerationInput_1a020cab6c3ca7089cd3462916cddb73a0"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">GenericGenerationInput</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">endId</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">padId</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput9TensorPtrE" title="tensorrt_llm::runtime::GenericGenerationInput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">ids</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput9TensorPtrE" title="tensorrt_llm::runtime::GenericGenerationInput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">lengths</span></span>, <span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">packed</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">false</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput22GenericGenerationInputEK10SizeType32K10SizeType329TensorPtr9TensorPtrb" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput5endIdE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime22GenericGenerationInput5endIdE"></span><span id="_CPPv2N12tensorrt_llm7runtime22GenericGenerationInput5endIdE"></span><span id="tensorrt_llm::runtime::GenericGenerationInput::endId__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericGenerationInput_1aeb034308719f1b0dd64d60eb36a681e2"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">endId</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput5endIdE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput5padIdE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime22GenericGenerationInput5padIdE"></span><span id="_CPPv2N12tensorrt_llm7runtime22GenericGenerationInput5padIdE"></span><span id="tensorrt_llm::runtime::GenericGenerationInput::padId__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericGenerationInput_1a7ef9ea96e18bbc7941a72806c6a28d8a"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">padId</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput5padIdE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput3idsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime22GenericGenerationInput3idsE"></span><span id="_CPPv2N12tensorrt_llm7runtime22GenericGenerationInput3idsE"></span><span id="tensorrt_llm::runtime::GenericGenerationInput::ids__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericGenerationInput_1a33a188aa66efd03ca14459152178d6dd"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput9TensorPtrE" title="tensorrt_llm::runtime::GenericGenerationInput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">ids</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput3idsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput7lengthsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime22GenericGenerationInput7lengthsE"></span><span id="_CPPv2N12tensorrt_llm7runtime22GenericGenerationInput7lengthsE"></span><span id="tensorrt_llm::runtime::GenericGenerationInput::lengths__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericGenerationInput_1a6e25b8ca6502844fa99acc5ed8cb4cf0"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput9TensorPtrE" title="tensorrt_llm::runtime::GenericGenerationInput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">lengths</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput7lengthsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput6packedE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime22GenericGenerationInput6packedE"></span><span id="_CPPv2N12tensorrt_llm7runtime22GenericGenerationInput6packedE"></span><span id="tensorrt_llm::runtime::GenericGenerationInput::packed__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericGenerationInput_1a1251ffbef413840f81dc59bdedb88ee1"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">packed</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput6packedE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput13embeddingBiasE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime22GenericGenerationInput13embeddingBiasE"></span><span id="_CPPv2N12tensorrt_llm7runtime22GenericGenerationInput13embeddingBiasE"></span><span id="tensorrt_llm::runtime::GenericGenerationInput::embeddingBias__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericGenerationInput_1a69f8b8593ce1a89fd78f907b4c17cf55"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput9TensorPtrE" title="tensorrt_llm::runtime::GenericGenerationInput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">embeddingBias</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput13embeddingBiasE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput12badWordsListE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime22GenericGenerationInput12badWordsListE"></span><span id="_CPPv2N12tensorrt_llm7runtime22GenericGenerationInput12badWordsListE"></span><span id="tensorrt_llm::runtime::GenericGenerationInput::badWordsList__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericGenerationInput_1a7fdba88ec2c4ef3a749bf6730a8427f4"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput9TensorPtrE" title="tensorrt_llm::runtime::GenericGenerationInput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">badWordsList</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput12badWordsListE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput13stopWordsListE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime22GenericGenerationInput13stopWordsListE"></span><span id="_CPPv2N12tensorrt_llm7runtime22GenericGenerationInput13stopWordsListE"></span><span id="tensorrt_llm::runtime::GenericGenerationInput::stopWordsList__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericGenerationInput_1a13e840f74710c9ac836c64797652ebbe"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput9TensorPtrE" title="tensorrt_llm::runtime::GenericGenerationInput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">stopWordsList</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput13stopWordsListE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput12maxNewTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime22GenericGenerationInput12maxNewTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime22GenericGenerationInput12maxNewTokensE"></span><span id="tensorrt_llm::runtime::GenericGenerationInput::maxNewTokens__std::optional:SizeType32:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericGenerationInput_1aa2f7940f98cf9cc5650c4b423cbe47a7"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">maxNewTokens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput12maxNewTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput18promptTuningParamsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime22GenericGenerationInput18promptTuningParamsE"></span><span id="_CPPv2N12tensorrt_llm7runtime22GenericGenerationInput18promptTuningParamsE"></span><span id="tensorrt_llm::runtime::GenericGenerationInput::promptTuningParams__PromptTuningParams"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericGenerationInput_1a4da27c7a8e7b4bdd90c87c4bf9a08b46"></span><a class="reference internal" href="#_CPPv4I00EN12tensorrt_llm7runtime22GenericGenerationInputE" title="tensorrt_llm::runtime::GenericGenerationInput::PromptTuningParams"><span class="n"><span class="pre">PromptTuningParams</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">promptTuningParams</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime22GenericGenerationInput18promptTuningParamsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="generationoutput-h">
|
||
<h2>generationOutput.h<a class="headerlink" href="#generationoutput-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime16GenerationOutputE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16GenerationOutputE"></span><span id="_CPPv2N12tensorrt_llm7runtime16GenerationOutputE"></span><span id="tensorrt_llm::runtime::GenerationOutput"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenerationOutput"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">GenerationOutput</span></span></span><span class="w"> </span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="k"><span class="pre">public</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime23GenericGenerationOutputE" title="tensorrt_llm::runtime::GenericGenerationOutput"><span class="n"><span class="pre">GenericGenerationOutput</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime16GenerationOutputE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime16GenerationOutput4BaseE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16GenerationOutput4BaseE"></span><span id="_CPPv2N12tensorrt_llm7runtime16GenerationOutput4BaseE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenerationOutput_1ad51e9ff641b7853ab71d1aa3d584bc68"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Base</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime23GenericGenerationOutputE" title="tensorrt_llm::runtime::GenericGenerationOutput"><span class="n"><span class="pre">GenericGenerationOutput</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime16GenerationOutput4BaseE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime16GenerationOutput9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16GenerationOutput9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime16GenerationOutput9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenerationOutput_1a5c7e8fe9eade02e97e6576ecfe16b0df"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16GenerationOutput4BaseE" title="tensorrt_llm::runtime::GenerationOutput::Base"><span class="n"><span class="pre">Base</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">TensorPtr</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime16GenerationOutput9TensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime16GenerationOutput16GenerationOutputE9TensorPtr9TensorPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16GenerationOutput16GenerationOutputE9TensorPtr9TensorPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime16GenerationOutput16GenerationOutputE9TensorPtr9TensorPtr"></span><span id="tensorrt_llm::runtime::GenerationOutput::GenerationOutput__TensorPtr.TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenerationOutput_1a0c21ff4190e449fe91a596650c8d2146"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">GenerationOutput</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16GenerationOutput9TensorPtrE" title="tensorrt_llm::runtime::GenerationOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">ids</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16GenerationOutput9TensorPtrE" title="tensorrt_llm::runtime::GenerationOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">lengths</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime16GenerationOutput16GenerationOutputE9TensorPtr9TensorPtr" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime23GenericGenerationOutputE">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime23GenericGenerationOutputE"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime23GenericGenerationOutputE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TTensor</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericGenerationOutput"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">GenericGenerationOutput</span></span></span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime23GenericGenerationOutputE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="docutils container">
|
||
<em>#include <generationOutput.h></em></div>
|
||
<p><em></em></p>
|
||
<p><ul class="simple">
|
||
<li><p><code class="docutils literal notranslate"><span class="pre">ids</span></code>, is a tensor that contains the output token IDs. Its shape is <code class="docutils literal notranslate"><span class="pre">[batchSize,</span> <span class="pre">beamWidth,</span> <span class="pre">maxSeqLength]</span></code> where <code class="docutils literal notranslate"><span class="pre">maxSeqLength</span></code> is the sum of <code class="docutils literal notranslate"><span class="pre">maxInputLength</span></code> and <code class="docutils literal notranslate"><span class="pre">maxNewTokens</span></code>. After generation, it contains, for each sequence, a copy of the input tokens followed by the output tokens. When a sequence is shorter than <code class="docutils literal notranslate"><span class="pre">maxSeqLength</span></code>, padding tokens are added at the end of the sequence.</p></li>
|
||
</ul>
|
||
</p>
|
||
<p><em>Note that the shape of that tensor is different in this version of TensorRT-LLM from its shape in previous versions where it was </em>.</p>
|
||
<p><em></em></p>
|
||
<p><ul>
|
||
<li><p><code class="docutils literal notranslate"><span class="pre">logProbs</span></code>, is a tensor of floating-point values on the GPU to store the log-prob of the generated tokens. Its shape is <code class="docutils literal notranslate"><span class="pre">[maxNewTokens,</span> <span class="pre">batchSize,</span> <span class="pre">beamWidth]</span></code>. Its shape will likely change in a future release to match the shape of the output <code class="docutils literal notranslate"><span class="pre">ids</span></code> tensor.</p></li>
|
||
<li><p><code class="docutils literal notranslate"><span class="pre">contextLogits</span></code>, is a tensor of values on the GPU (same datatype as the computation type) to store the logits for the context. Its shape is <code class="docutils literal notranslate"><span class="pre">[batchSize,</span> <span class="pre">maxSequenceLength,</span> <span class="pre">vocabSizePadded]</span></code>. If use <code class="docutils literal notranslate"><span class="pre">remove_input_padding</span></code>, its shape is <code class="docutils literal notranslate"><span class="pre">[packedSize,</span> <span class="pre">vocabSizePadded]</span></code>. This buffer will only be filled in if the TensorRT engine was built with the <code class="docutils literal notranslate"><span class="pre">gather_context_logits</span></code> or <code class="docutils literal notranslate"><span class="pre">gather_all_token_logits</span></code> parameter enabled.</p>
|
||
<p>After inference is complete, you can get the context logits in <code class="docutils literal notranslate"><span class="pre">GenerationOutput.contextLogits</span></code>, these are variables on the GPU. For specific acquisition methods, please refer to the example of <a class="reference external" href="https://github.com/NVIDIA/TensorRT-LLM/blob/main/benchmarks/cpp/gptSessionBenchmark.cpp">gptSessionBenchmark.cpp</a>.</p>
|
||
<p>It is important to point out that enabling the computation may have an impact on performance (the language modeling head (LM head) has to perform a matrix multiplication on all the context tokens instead of a just the last one).</p>
|
||
</li>
|
||
<li><p><code class="docutils literal notranslate"><span class="pre">generationLogits</span></code>, is a tensor of values on the GPU (same datatype as the computation type) to store the logits for the generation. Its shape is <code class="docutils literal notranslate"><span class="pre">[batchSize,</span> <span class="pre">beamWidth,</span> <span class="pre">maxOutputLen,</span> <span class="pre">vocabSizePadded]</span></code>. This buffer will only be filled in if the TensorRT engine was built with the <code class="docutils literal notranslate"><span class="pre">gather_generation_logits</span></code> or <code class="docutils literal notranslate"><span class="pre">gather_all_token_logits</span></code> parameter enabled.</p>
|
||
<p>Generation logits can also be obtained through <code class="docutils literal notranslate"><span class="pre">GenerationOutput.generationLogits</span></code> after inference is completed.</p>
|
||
</li>
|
||
<li><p><code class="docutils literal notranslate"><span class="pre">onTokenGenerated</span></code>, is a callback function invoked in the generation loop to pass newly generated tokens to the caller while the loop continues to execute. An implementation of that callback must accept the output <code class="docutils literal notranslate"><span class="pre">ids</span></code> tensor, the generation <code class="docutils literal notranslate"><span class="pre">step</span></code> and a boolean flag that indicates if the generation is complete. </p></li>
|
||
</ul>
|
||
</p>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23GenericGenerationOutput9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime23GenericGenerationOutput9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericGenerationOutput_1a739b5cbc24692f24b77a467612857e35"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime23GenericGenerationOutputE" title="tensorrt_llm::runtime::GenericGenerationOutput::TTensor"><span class="n"><span class="pre">TTensor</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput9TensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput8CallbackE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23GenericGenerationOutput8CallbackE"></span><span id="_CPPv2N12tensorrt_llm7runtime23GenericGenerationOutput8CallbackE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericGenerationOutput_1a32e3b1365c31a57820038c5a6ae4ff30"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Callback</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">function</span></span><span class="p"><span class="pre"><</span></span><span class="kt"><span class="pre">void</span></span><span class="p"><span class="pre">(</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput9TensorPtrE" title="tensorrt_llm::runtime::GenericGenerationOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n"><span class="pre">ids</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n"><span class="pre">step</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n"><span class="pre">finished</span></span><span class="p"><span class="pre">)</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput8CallbackE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput23GenericGenerationOutputE9TensorPtr9TensorPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23GenericGenerationOutput23GenericGenerationOutputE9TensorPtr9TensorPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime23GenericGenerationOutput23GenericGenerationOutputE9TensorPtr9TensorPtr"></span><span id="tensorrt_llm::runtime::GenericGenerationOutput::GenericGenerationOutput__TensorPtr.TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericGenerationOutput_1a3ed16459b1913958c761070564be3754"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">GenericGenerationOutput</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput9TensorPtrE" title="tensorrt_llm::runtime::GenericGenerationOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">ids</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput9TensorPtrE" title="tensorrt_llm::runtime::GenericGenerationOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">lengths</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput23GenericGenerationOutputE9TensorPtr9TensorPtr" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput3idsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23GenericGenerationOutput3idsE"></span><span id="_CPPv2N12tensorrt_llm7runtime23GenericGenerationOutput3idsE"></span><span id="tensorrt_llm::runtime::GenericGenerationOutput::ids__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericGenerationOutput_1a7ff8e99b46a473dbfc53aecc127449bd"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput9TensorPtrE" title="tensorrt_llm::runtime::GenericGenerationOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">ids</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput3idsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput7lengthsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23GenericGenerationOutput7lengthsE"></span><span id="_CPPv2N12tensorrt_llm7runtime23GenericGenerationOutput7lengthsE"></span><span id="tensorrt_llm::runtime::GenericGenerationOutput::lengths__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericGenerationOutput_1a53bb028de3dd943d7d0913753e522aae"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput9TensorPtrE" title="tensorrt_llm::runtime::GenericGenerationOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">lengths</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput7lengthsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput11cumLogProbsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23GenericGenerationOutput11cumLogProbsE"></span><span id="_CPPv2N12tensorrt_llm7runtime23GenericGenerationOutput11cumLogProbsE"></span><span id="tensorrt_llm::runtime::GenericGenerationOutput::cumLogProbs__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericGenerationOutput_1a3f7b1e14ccdaa0c7525f90d32e1c43d6"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput9TensorPtrE" title="tensorrt_llm::runtime::GenericGenerationOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">cumLogProbs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput11cumLogProbsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput8logProbsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23GenericGenerationOutput8logProbsE"></span><span id="_CPPv2N12tensorrt_llm7runtime23GenericGenerationOutput8logProbsE"></span><span id="tensorrt_llm::runtime::GenericGenerationOutput::logProbs__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericGenerationOutput_1ae28cb59770ea150c435484b4f7f453d3"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput9TensorPtrE" title="tensorrt_llm::runtime::GenericGenerationOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">logProbs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput8logProbsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput13contextLogitsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23GenericGenerationOutput13contextLogitsE"></span><span id="_CPPv2N12tensorrt_llm7runtime23GenericGenerationOutput13contextLogitsE"></span><span id="tensorrt_llm::runtime::GenericGenerationOutput::contextLogits__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericGenerationOutput_1a884e4fabdeb57a635f796ddc36274abc"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput9TensorPtrE" title="tensorrt_llm::runtime::GenericGenerationOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">contextLogits</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput13contextLogitsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput16generationLogitsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23GenericGenerationOutput16generationLogitsE"></span><span id="_CPPv2N12tensorrt_llm7runtime23GenericGenerationOutput16generationLogitsE"></span><span id="tensorrt_llm::runtime::GenericGenerationOutput::generationLogits__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericGenerationOutput_1ad76f135de78a3ab89804cbf3f7175d0f"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput9TensorPtrE" title="tensorrt_llm::runtime::GenericGenerationOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">generationLogits</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput16generationLogitsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput16onTokenGeneratedE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23GenericGenerationOutput16onTokenGeneratedE"></span><span id="_CPPv2N12tensorrt_llm7runtime23GenericGenerationOutput16onTokenGeneratedE"></span><span id="tensorrt_llm::runtime::GenericGenerationOutput::onTokenGenerated__Callback"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericGenerationOutput_1a35249ba2dc1990149ddc5895c2e3a1f7"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput8CallbackE" title="tensorrt_llm::runtime::GenericGenerationOutput::Callback"><span class="n"><span class="pre">Callback</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">onTokenGenerated</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23GenericGenerationOutput16onTokenGeneratedE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="gptjsonconfig-h">
|
||
<h2>gptJsonConfig.h<a class="headerlink" href="#gptjsonconfig-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13GptJsonConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13GptJsonConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime13GptJsonConfigE"></span><span id="tensorrt_llm::runtime::GptJsonConfig"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">GptJsonConfig</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13GptJsonConfig13GptJsonConfigENSt6stringENSt6stringENSt6stringE10SizeType3210SizeType3210SizeType3210SizeType3211ModelConfigNSt8optionalI15RuntimeDefaultsEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13GptJsonConfig13GptJsonConfigENSt6stringENSt6stringENSt6stringE10SizeType3210SizeType3210SizeType3210SizeType3211ModelConfigNSt8optionalI15RuntimeDefaultsEE"></span><span id="_CPPv2N12tensorrt_llm7runtime13GptJsonConfig13GptJsonConfigENSt6stringENSt6stringENSt6stringE10SizeType3210SizeType3210SizeType3210SizeType3211ModelConfigNSt8optionalI15RuntimeDefaultsEE"></span><span id="tensorrt_llm::runtime::GptJsonConfig::GptJsonConfig__ss.ss.ss.SizeType32.SizeType32.SizeType32.SizeType32.ModelConfig.std::optional:RuntimeDefaults:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1a712400cdaee8fc97ce35e2299ab1f4af"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">GptJsonConfig</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">name</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">version</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">precision</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tensorParallelism</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">pipelineParallelism</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">contextParallelism</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">gpusPerNode</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE" title="tensorrt_llm::runtime::ModelConfig"><span class="n"><span class="pre">ModelConfig</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">modelConfig</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15RuntimeDefaultsE" title="tensorrt_llm::runtime::RuntimeDefaults"><span class="n"><span class="pre">RuntimeDefaults</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="n sig-param"><span class="pre">runtimeDefaults</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">nullopt</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig13GptJsonConfigENSt6stringENSt6stringENSt6stringE10SizeType3210SizeType3210SizeType3210SizeType3211ModelConfigNSt8optionalI15RuntimeDefaultsEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig14getModelConfigEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13GptJsonConfig14getModelConfigEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime13GptJsonConfig14getModelConfigEv"></span><span id="tensorrt_llm::runtime::GptJsonConfig::getModelConfigC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1a5c3285f2c061c1330ca11e1343cf89a2"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE" title="tensorrt_llm::runtime::ModelConfig"><span class="n"><span class="pre">ModelConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="n"><span class="pre">getModelConfig</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig14getModelConfigEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13GptJsonConfig21getModelConfigMutableEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13GptJsonConfig21getModelConfigMutableEv"></span><span id="_CPPv2N12tensorrt_llm7runtime13GptJsonConfig21getModelConfigMutableEv"></span><span id="tensorrt_llm::runtime::GptJsonConfig::getModelConfigMutable"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1ad6d1468e250b06a17e67e10b7d94d823"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE" title="tensorrt_llm::runtime::ModelConfig"><span class="n"><span class="pre">ModelConfig</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="n"><span class="pre">getModelConfigMutable</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig21getModelConfigMutableEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig7getNameEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13GptJsonConfig7getNameEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime13GptJsonConfig7getNameEv"></span><span id="tensorrt_llm::runtime::GptJsonConfig::getNameC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1a5270fbe10703e1e5fe5d52d104b30cfb"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="n"><span class="pre">getName</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig7getNameEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig10getVersionEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13GptJsonConfig10getVersionEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime13GptJsonConfig10getVersionEv"></span><span id="tensorrt_llm::runtime::GptJsonConfig::getVersionC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1a5900e7d368979420ec10c7635e656ea3"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="n"><span class="pre">getVersion</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig10getVersionEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig12getPrecisionEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13GptJsonConfig12getPrecisionEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime13GptJsonConfig12getPrecisionEv"></span><span id="tensorrt_llm::runtime::GptJsonConfig::getPrecisionC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1a353424dd7b526c63e5d03fedc5314ed1"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="n"><span class="pre">getPrecision</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig12getPrecisionEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig20getTensorParallelismEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13GptJsonConfig20getTensorParallelismEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime13GptJsonConfig20getTensorParallelismEv"></span><span id="tensorrt_llm::runtime::GptJsonConfig::getTensorParallelismCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1aac7f508e90ae3316d1065b0eb68f8aee"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getTensorParallelism</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig20getTensorParallelismEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig22getPipelineParallelismEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13GptJsonConfig22getPipelineParallelismEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime13GptJsonConfig22getPipelineParallelismEv"></span><span id="tensorrt_llm::runtime::GptJsonConfig::getPipelineParallelismCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1a075d679a1b8c80bc303441308a9c28bc"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getPipelineParallelism</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig22getPipelineParallelismEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig21getContextParallelismEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13GptJsonConfig21getContextParallelismEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime13GptJsonConfig21getContextParallelismEv"></span><span id="tensorrt_llm::runtime::GptJsonConfig::getContextParallelismCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1ad441fbfc413a64d424510520526a38b7"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getContextParallelism</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig21getContextParallelismEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig14getGpusPerNodeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13GptJsonConfig14getGpusPerNodeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime13GptJsonConfig14getGpusPerNodeEv"></span><span id="tensorrt_llm::runtime::GptJsonConfig::getGpusPerNodeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1ad3b5423d2e378a551e73abce93609667"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getGpusPerNode</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig14getGpusPerNodeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig12getWorldSizeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13GptJsonConfig12getWorldSizeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime13GptJsonConfig12getWorldSizeEv"></span><span id="tensorrt_llm::runtime::GptJsonConfig::getWorldSizeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1a3023e339b22eca5cf4ba14304c46b567"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getWorldSize</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig12getWorldSizeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig18getRuntimeDefaultsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13GptJsonConfig18getRuntimeDefaultsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime13GptJsonConfig18getRuntimeDefaultsEv"></span><span id="tensorrt_llm::runtime::GptJsonConfig::getRuntimeDefaultsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1ac8e7e4aa82a1c14146563b5ee7acd7b8"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15RuntimeDefaultsE" title="tensorrt_llm::runtime::RuntimeDefaults"><span class="n"><span class="pre">RuntimeDefaults</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getRuntimeDefaults</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig18getRuntimeDefaultsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig14engineFilenameERK11WorldConfigRKNSt6stringE">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13GptJsonConfig14engineFilenameERK11WorldConfigRKNSt6stringE"></span><span id="_CPPv2NK12tensorrt_llm7runtime13GptJsonConfig14engineFilenameERK11WorldConfigRKNSt6stringE"></span><span id="tensorrt_llm::runtime::GptJsonConfig::engineFilename__WorldConfigCR.ssCRC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1ac663861699ef7f573f64d5fdb89f14af"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">engineFilename</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfigE" title="tensorrt_llm::runtime::WorldConfig"><span class="n"><span class="pre">WorldConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">worldConfig</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">model</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig14engineFilenameERK11WorldConfigRKNSt6stringE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig14engineFilenameERK11WorldConfig">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime13GptJsonConfig14engineFilenameERK11WorldConfig"></span><span id="_CPPv2NK12tensorrt_llm7runtime13GptJsonConfig14engineFilenameERK11WorldConfig"></span><span id="tensorrt_llm::runtime::GptJsonConfig::engineFilename__WorldConfigCRC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1a9ce08952a809fbe5859a685215dd7258"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">engineFilename</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfigE" title="tensorrt_llm::runtime::WorldConfig"><span class="n"><span class="pre">WorldConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">worldConfig</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime13GptJsonConfig14engineFilenameERK11WorldConfig" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-functions">Public Static Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13GptJsonConfig5parseERKNSt6stringE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13GptJsonConfig5parseERKNSt6stringE"></span><span id="_CPPv2N12tensorrt_llm7runtime13GptJsonConfig5parseERKNSt6stringE"></span><span id="tensorrt_llm::runtime::GptJsonConfig::parse__ssCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1a258c808e0e2c5445c807ae2f0a257782"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfigE" title="tensorrt_llm::runtime::GptJsonConfig"><span class="n"><span class="pre">GptJsonConfig</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">parse</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">json</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig5parseERKNSt6stringE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13GptJsonConfig5parseERNSt7istreamE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13GptJsonConfig5parseERNSt7istreamE"></span><span id="_CPPv2N12tensorrt_llm7runtime13GptJsonConfig5parseERNSt7istreamE"></span><span id="tensorrt_llm::runtime::GptJsonConfig::parse__isR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1af74ea4db9854b1f0ddd8aaa0d4a48f39"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfigE" title="tensorrt_llm::runtime::GptJsonConfig"><span class="n"><span class="pre">GptJsonConfig</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">parse</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">istream</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">json</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig5parseERNSt7istreamE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13GptJsonConfig5parseERKNSt10filesystem4pathE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13GptJsonConfig5parseERKNSt10filesystem4pathE"></span><span id="_CPPv2N12tensorrt_llm7runtime13GptJsonConfig5parseERKNSt10filesystem4pathE"></span><span id="tensorrt_llm::runtime::GptJsonConfig::parse__std::filesystem::pathCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1a959c20c852124ced898b12bbdeb5d79f"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfigE" title="tensorrt_llm::runtime::GptJsonConfig"><span class="n"><span class="pre">GptJsonConfig</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">parse</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">filesystem</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">path</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">path</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig5parseERKNSt10filesystem4pathE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-members">Private Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13GptJsonConfig5mNameE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13GptJsonConfig5mNameE"></span><span id="_CPPv2N12tensorrt_llm7runtime13GptJsonConfig5mNameE"></span><span id="tensorrt_llm::runtime::GptJsonConfig::mName__ssC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1aa23cc7f0c9ad465ffdb3c1950876b470"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mName</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig5mNameE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13GptJsonConfig8mVersionE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13GptJsonConfig8mVersionE"></span><span id="_CPPv2N12tensorrt_llm7runtime13GptJsonConfig8mVersionE"></span><span id="tensorrt_llm::runtime::GptJsonConfig::mVersion__ssC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1a254b88b5fe490e3835177dd32ecb2c8f"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mVersion</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig8mVersionE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13GptJsonConfig10mPrecisionE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13GptJsonConfig10mPrecisionE"></span><span id="_CPPv2N12tensorrt_llm7runtime13GptJsonConfig10mPrecisionE"></span><span id="tensorrt_llm::runtime::GptJsonConfig::mPrecision__ssC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1acf01c9bba53cd18d0ece1026ef19e485"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mPrecision</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig10mPrecisionE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13GptJsonConfig18mTensorParallelismE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13GptJsonConfig18mTensorParallelismE"></span><span id="_CPPv2N12tensorrt_llm7runtime13GptJsonConfig18mTensorParallelismE"></span><span id="tensorrt_llm::runtime::GptJsonConfig::mTensorParallelism__SizeType32C"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1ae1cc00036f7255a4d9580f833f8ab146"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mTensorParallelism</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig18mTensorParallelismE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13GptJsonConfig20mPipelineParallelismE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13GptJsonConfig20mPipelineParallelismE"></span><span id="_CPPv2N12tensorrt_llm7runtime13GptJsonConfig20mPipelineParallelismE"></span><span id="tensorrt_llm::runtime::GptJsonConfig::mPipelineParallelism__SizeType32C"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1a5eef99e58f6d89bb63040884a26f4dd6"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mPipelineParallelism</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig20mPipelineParallelismE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13GptJsonConfig19mContextParallelismE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13GptJsonConfig19mContextParallelismE"></span><span id="_CPPv2N12tensorrt_llm7runtime13GptJsonConfig19mContextParallelismE"></span><span id="tensorrt_llm::runtime::GptJsonConfig::mContextParallelism__SizeType32C"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1a401239ff678e6761f2ab44f2da2e8480"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mContextParallelism</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig19mContextParallelismE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13GptJsonConfig12mGpusPerNodeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13GptJsonConfig12mGpusPerNodeE"></span><span id="_CPPv2N12tensorrt_llm7runtime13GptJsonConfig12mGpusPerNodeE"></span><span id="tensorrt_llm::runtime::GptJsonConfig::mGpusPerNode__SizeType32C"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1a719eb6ee79619bdf2997de06d7d93d52"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mGpusPerNode</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig12mGpusPerNodeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13GptJsonConfig12mModelConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13GptJsonConfig12mModelConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime13GptJsonConfig12mModelConfigE"></span><span id="tensorrt_llm::runtime::GptJsonConfig::mModelConfig__ModelConfig"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1ada6b440f8e35d7cc663ae5aee2202655"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE" title="tensorrt_llm::runtime::ModelConfig"><span class="n"><span class="pre">ModelConfig</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mModelConfig</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig12mModelConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13GptJsonConfig16mRuntimeDefaultsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13GptJsonConfig16mRuntimeDefaultsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13GptJsonConfig16mRuntimeDefaultsE"></span><span id="tensorrt_llm::runtime::GptJsonConfig::mRuntimeDefaults__std::optional:RuntimeDefaults:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1abb7f59b5946ca00c70649e7c2554028a"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15RuntimeDefaultsE" title="tensorrt_llm::runtime::RuntimeDefaults"><span class="n"><span class="pre">RuntimeDefaults</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mRuntimeDefaults</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig16mRuntimeDefaultsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="itensor-h">
|
||
<h2>iTensor.h<a class="headerlink" href="#itensor-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv48nvinfer1">
|
||
<span id="_CPPv38nvinfer1"></span><span id="_CPPv28nvinfer1"></span><span id="nvinfer1"></span><span class="target" id="namespacenvinfer1"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">nvinfer1</span></span></span><a class="headerlink" href="#_CPPv48nvinfer1" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-functions">Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtimelsERNSt7ostreamERKN7ITensor5ShapeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtimelsERNSt7ostreamERKN7ITensor5ShapeE"></span><span id="_CPPv2N12tensorrt_llm7runtimelsERNSt7ostreamERKN7ITensor5ShapeE"></span><span id="tensorrt_llm::runtime::lshift-operator__osR.ITensor::ShapeCR"></span><span class="target" id="iTensor_8h_1a0c96496836cd2bdc559321f5e7c31294"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ostream</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="k"><span class="pre">operator</span></span><span class="o"><span class="pre"><<</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ostream</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">output</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">dims</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtimelsERNSt7ostreamERKN7ITensor5ShapeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Utility function to print a shape. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtimelsERNSt7ostreamERK7ITensor">
|
||
<span id="_CPPv3N12tensorrt_llm7runtimelsERNSt7ostreamERK7ITensor"></span><span id="_CPPv2N12tensorrt_llm7runtimelsERNSt7ostreamERK7ITensor"></span><span id="tensorrt_llm::runtime::lshift-operator__osR.ITensorCR"></span><span class="target" id="iTensor_8h_1a9169197d73d3cf134f280862c275af26"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ostream</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="k"><span class="pre">operator</span></span><span class="o"><span class="pre"><<</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ostream</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">output</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">tensor</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtimelsERNSt7ostreamERK7ITensor" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Utility function to print a tensor with its shape. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEPK1TRKN7ITensor14SharedConstPtrE">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime16bufferCastOrNullERKN7ITensor14SharedConstPtrE"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime16bufferCastOrNullERKN7ITensor14SharedConstPtrE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="iTensor_8h_1a52b485dcf9191f3234857aea11a34e09"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEPK1TRKN7ITensor14SharedConstPtrE" title="tensorrt_llm::runtime::bufferCastOrNull::T"><span class="n"><span class="pre">T</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="sig-name descname"><span class="n"><span class="pre">bufferCastOrNull</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor14SharedConstPtrE" title="tensorrt_llm::runtime::ITensor::SharedConstPtr"><span class="n"><span class="pre">SharedConstPtr</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">tensorPtr</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEPK1TRKN7ITensor14SharedConstPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Retrieves a T const typed pointer to the underlying data of the tensor pointed to by the tensorPtr, or nullptr if the tensorPtr is null. </p>
|
||
<p>This overload has to be declared to avoid ambiguity when an implicit conversion to <a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1IBuffer"><span class="std std-ref">IBuffer</span></a> is involved. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Template Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>T</strong> – The type of the underlying data. </p>
|
||
</dd>
|
||
<dt class="field-even">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p><strong>tensorPtr</strong> – A possibly null shared ptr. </p>
|
||
</dd>
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>A pointer to T const, possibly nullptr. </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEP1TRKN7ITensor9SharedPtrE">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime16bufferCastOrNullERKN7ITensor9SharedPtrE"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime16bufferCastOrNullERKN7ITensor9SharedPtrE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="iTensor_8h_1a7167b962238a8d59bcd7187f3c35595b"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEP1TRKN7ITensor9SharedPtrE" title="tensorrt_llm::runtime::bufferCastOrNull::T"><span class="n"><span class="pre">T</span></span></a><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="sig-name descname"><span class="n"><span class="pre">bufferCastOrNull</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">tensorPtr</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEP1TRKN7ITensor9SharedPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Retrieves a T typed pointer to the underlying data of the buffer pointed to by the tensorPtr, or nullptr if the tensorPtr is null. </p>
|
||
<p>This overload has to be declared to avoid ambiguity when an implicit conversion to <a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1IBuffer"><span class="std std-ref">IBuffer</span></a> is involved. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Template Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>T</strong> – The type of the underlying data. </p>
|
||
</dd>
|
||
<dt class="field-even">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p><strong>tensorPtr</strong> – A possibly null shared ptr. </p>
|
||
</dd>
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>A pointer to T, possibly nullptr. </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEP1TRKNSt8optionalIN7ITensor9SharedPtrEEE">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime16bufferCastOrNullERKNSt8optionalIN7ITensor9SharedPtrEEE"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime16bufferCastOrNullERKNSt8optionalIN7ITensor9SharedPtrEEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="iTensor_8h_1ad9a080040e1bda831a207038efb156d7"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEP1TRKNSt8optionalIN7ITensor9SharedPtrEEE" title="tensorrt_llm::runtime::bufferCastOrNull::T"><span class="n"><span class="pre">T</span></span></a><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="sig-name descname"><span class="n"><span class="pre">bufferCastOrNull</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">optionalTensorPtr</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEP1TRKNSt8optionalIN7ITensor9SharedPtrEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Retrieves a T typed pointer to the underlying data of the tensor pointed to by the tensor pointer contained in the optionalBufferPtr, or nullptr if the optional doesn’t have a value. </p>
|
||
<p>This overload has to be declared to avoid ambiguity when an implicit conversion to <a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1IBuffer"><span class="std std-ref">IBuffer</span></a> is involved. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Template Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>T</strong> – The type of the underlying data. </p>
|
||
</dd>
|
||
<dt class="field-even">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p><strong>optionalBufferPtr</strong> – A possibly empty optional. </p>
|
||
</dd>
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>A pointer to T, possibly nullptr. </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEPK1TRKNSt8optionalIN7ITensor14SharedConstPtrEEE">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime16bufferCastOrNullERKNSt8optionalIN7ITensor14SharedConstPtrEEE"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime16bufferCastOrNullERKNSt8optionalIN7ITensor14SharedConstPtrEEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="iTensor_8h_1afc77303772dfea61ffd36f5fa06e73e6"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEPK1TRKNSt8optionalIN7ITensor14SharedConstPtrEEE" title="tensorrt_llm::runtime::bufferCastOrNull::T"><span class="n"><span class="pre">T</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="sig-name descname"><span class="n"><span class="pre">bufferCastOrNull</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor14SharedConstPtrE" title="tensorrt_llm::runtime::ITensor::SharedConstPtr"><span class="n"><span class="pre">SharedConstPtr</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">optionalTensorPtr</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEPK1TRKNSt8optionalIN7ITensor14SharedConstPtrEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Retrieves a T const typed pointer to the underlying data of the tensor pointed to by the tensor pointer contained in the optionalBufferPtr, or nullptr if the optional doesn’t have a value. </p>
|
||
<p>This overload has to be declared to avoid ambiguity when an implicit conversion to <a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1IBuffer"><span class="std std-ref">IBuffer</span></a> is involved. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Template Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>T</strong> – The type of the underlying data. </p>
|
||
</dd>
|
||
<dt class="field-even">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p><strong>optionalBufferPtr</strong> – A possibly empty optional. </p>
|
||
</dd>
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>A pointer to const T, possibly nullptr. </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
</div>
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensorE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensorE"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensorE"></span><span id="tensorrt_llm::runtime::ITensor"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">ITensor</span></span></span><span class="w"> </span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="k"><span class="pre">public</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor9UniquePtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor9UniquePtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor9UniquePtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a0ebcd2ee37abad2960daf1346816f348"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">UniquePtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">unique_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9UniquePtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor9SharedPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor9SharedPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a31e67140fadaeb6eaaf096ff37d640ad"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">SharedPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor14UniqueConstPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor14UniqueConstPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor14UniqueConstPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1ad199f95dd25dfbef298cc7d1f3655d66"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">UniqueConstPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">unique_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor14UniqueConstPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor14SharedConstPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor14SharedConstPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor14SharedConstPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a591e6b0aa01a0043967262ffdcc3974d"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">SharedConstPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor14SharedConstPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor5ShapeE"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor5ShapeE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1ab2e78d36e5cc72b4443e832a46147fcc"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Shape</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">Dims</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor9DimType64E">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor9DimType64E"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor9DimType64E"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a0e1b3d629571428c37c511b70a05b334"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">DimType64</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">remove_reference_t</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">decltype</span></span><span class="p"><span class="pre">(</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">d</span></span><span class="p"><span class="pre">[</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">]</span></span><span class="p"><span class="pre">)</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9DimType64E" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor9TensorMapE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor9TensorMapE"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor9TensorMapE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1ada89b277ed3347bbb527e3092a8eab04"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorMap</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime12StringPtrMapE" title="tensorrt_llm::runtime::StringPtrMap"><span class="n"><span class="pre">StringPtrMap</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9TensorMapE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensorD0Ev">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensorD0Ev"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensorD0Ev"></span><span id="tensorrt_llm::runtime::ITensor::~ITensor"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a857da95b81330009b368308ed0c2d3d9"></span><span class="sig-name descname"><span class="n"><span class="pre">~ITensor</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">override</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">default</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensorD0Ev" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime7ITensor8getShapeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime7ITensor8getShapeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime7ITensor8getShapeEv"></span><span id="tensorrt_llm::runtime::ITensor::getShapeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a8efdcf81157ec42fe414411891370355"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="n"><span class="pre">getShape</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime7ITensor8getShapeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Returns the tensor dimensions. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I_10SizeType32ENK12tensorrt_llm7runtime7ITensor12getDimensionE9DimType64v">
|
||
<span id="_CPPv3I_10SizeType32ENK12tensorrt_llm7runtime7ITensor12getDimensionEv"></span><span id="_CPPv2I_10SizeType32ENK12tensorrt_llm7runtime7ITensor12getDimensionEv"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">n</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a9d9640edefab8053cd81ed3e96bedd32"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9DimType64E" title="tensorrt_llm::runtime::ITensor::DimType64"><span class="n"><span class="pre">DimType64</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getDimension</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4I_10SizeType32ENK12tensorrt_llm7runtime7ITensor12getDimensionE9DimType64v" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Returns the tensor n-th dimension. If n is negative, returns the (nbDims - n)th dimension. TODO: replace with constexpr parameter when moving to C++20. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor7reshapeERK5Shape">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor7reshapeERK5Shape"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor7reshapeERK5Shape"></span><span id="tensorrt_llm::runtime::ITensor::reshape__ShapeCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a772377264b089406fb1cf743ac3163f3"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">reshape</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">dims</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor7reshapeERK5Shape" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Sets the tensor dimensions. The new size of the tensor will be <code class="docutils literal notranslate"><span class="pre">volume(dims)</span></code></p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor6resizeENSt6size_tE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor6resizeENSt6size_tE"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor6resizeENSt6size_tE"></span><span id="tensorrt_llm::runtime::ITensor::resize__std::s"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a8ed7ae223a11d8d516216a4d28286f62"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">resize</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">newSize</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor6resizeENSt6size_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Resizes the buffer. This is a no-op if the new size is smaller than or equal to the current capacity. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor7ITensorERK7ITensor">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor7ITensorERK7ITensor"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor7ITensorERK7ITensor"></span><span id="tensorrt_llm::runtime::ITensor::ITensor__ITensorCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1ade4b1d2a214d378f73255357e32bee1f"></span><span class="sig-name descname"><span class="n"><span class="pre">ITensor</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor7ITensorERK7ITensor" title="tensorrt_llm::runtime::ITensor::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="p"><span class="pre">&</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">delete</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor7ITensorERK7ITensor" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Not allowed to copy. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensoraSERK7ITensor">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensoraSERK7ITensor"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensoraSERK7ITensor"></span><span id="tensorrt_llm::runtime::ITensor::assign-operator__ITensorCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1afa9496881fe6d311048891210d206a83"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="k"><span class="pre">operator</span></span><span class="o"><span class="pre">=</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="p"><span class="pre">&</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">delete</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensoraSERK7ITensor" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Not allowed to copy. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor7squeezeE10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor7squeezeE10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor7squeezeE10SizeType32"></span><span id="tensorrt_llm::runtime::ITensor::squeeze__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a19b6e0dda1badc00eee1c8d558a627bd"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">squeeze</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">dim</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor7squeezeE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Removes the given <em>unit</em> dimensions from this tensor. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor9unsqueezeE10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor9unsqueezeE10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor9unsqueezeE10SizeType32"></span><span id="tensorrt_llm::runtime::ITensor::unsqueeze__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a93cd46c1307565a785482a58a454f5a1"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">unsqueeze</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">dim</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9unsqueezeE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Adds a <em>unit</em> dimension at the specified position. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime7ITensor11shapeEqualsERK5Shape">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime7ITensor11shapeEqualsERK5Shape"></span><span id="_CPPv2NK12tensorrt_llm7runtime7ITensor11shapeEqualsERK5Shape"></span><span id="tensorrt_llm::runtime::ITensor::shapeEquals__ShapeCRC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1ab84847a01479731500363a637003e03a"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">shapeEquals</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">other</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime7ITensor11shapeEqualsERK5Shape" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime7ITensor11shapeEqualsERKNSt16initializer_listI10SizeType32EE">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime7ITensor11shapeEqualsERKNSt16initializer_listI10SizeType32EE"></span><span id="_CPPv2NK12tensorrt_llm7runtime7ITensor11shapeEqualsERKNSt16initializer_listI10SizeType32EE"></span><span id="tensorrt_llm::runtime::ITensor::shapeEquals__std::initializer_list:SizeType32:CRC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1acc94dcf55ed9f128a97581822d735239"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">shapeEquals</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">initializer_list</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">other</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime7ITensor11shapeEqualsERKNSt16initializer_listI10SizeType32EE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0ENK12tensorrt_llm7runtime7ITensor11shapeEqualsEbPK1T10SizeType32">
|
||
<span id="_CPPv3I0ENK12tensorrt_llm7runtime7ITensor11shapeEqualsEPK1T10SizeType32"></span><span id="_CPPv2I0ENK12tensorrt_llm7runtime7ITensor11shapeEqualsEPK1T10SizeType32"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a90290f101c7fccb5719d975820f861a0"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">shapeEquals</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4I0ENK12tensorrt_llm7runtime7ITensor11shapeEqualsEbPK1T10SizeType32" title="tensorrt_llm::runtime::ITensor::shapeEquals::T"><span class="n"><span class="pre">T</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="n sig-param"><span class="pre">dims</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">count</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4I0ENK12tensorrt_llm7runtime7ITensor11shapeEqualsEbPK1T10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-functions">Public Static Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor6volumeERK5Shape">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor6volumeERK5Shape"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor6volumeERK5Shape"></span><span id="tensorrt_llm::runtime::ITensor::volume__ShapeCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a2afffa26551b43982bf51fa4920d4ade"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">int64_t</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">volume</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">dims</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor6volumeERK5Shape" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Returns the volume of the dimensions. Returns -1 if <code class="docutils literal notranslate"><span class="pre">d.nbDims</span> <span class="pre"><</span> <span class="pre">0</span></code>. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor17volumeNonNegativeERK5Shape">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor17volumeNonNegativeERK5Shape"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor17volumeNonNegativeERK5Shape"></span><span id="tensorrt_llm::runtime::ITensor::volumeNonNegative__ShapeCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1aac34c9309ce068d8216a54343348e21d"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">volumeNonNegative</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">shape</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor17volumeNonNegativeERK5Shape" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Returns the volume of the dimensions. Throws if <code class="docutils literal notranslate"><span class="pre">d.nbDims</span> <span class="pre"><</span> <span class="pre">0</span></code>. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor7stridesERK5Shape">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor7stridesERK5Shape"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor7stridesERK5Shape"></span><span id="tensorrt_llm::runtime::ITensor::strides__ShapeCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a9a9b3844320d4e08705c892320ef526c"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">strides</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">dims</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor7stridesERK5Shape" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Returns the strides of each dimemsion in a Shape. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor7squeezeERK5Shape10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor7squeezeERK5Shape10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor7squeezeERK5Shape10SizeType32"></span><span id="tensorrt_llm::runtime::ITensor::squeeze__ShapeCR.SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1adb620fc3a8eeac8e8502f8e864e3d822"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">squeeze</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">shape</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">dim</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor7squeezeERK5Shape10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Removes the given <em>unit</em> dimension from <code class="docutils literal notranslate"><span class="pre">shape</span></code>. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><ul class="simple">
|
||
<li><p><strong>shape</strong> – The shape to squeeze. </p></li>
|
||
<li><p><strong>dim</strong> – The dimension that should be removed (“squeezed”). </p></li>
|
||
</ul>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>A new shape without the unit dimension. </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor9unsqueezeERK5Shape10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor9unsqueezeERK5Shape10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor9unsqueezeERK5Shape10SizeType32"></span><span id="tensorrt_llm::runtime::ITensor::unsqueeze__ShapeCR.SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1af34ca957b146fa051184160c1db78af0"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">unsqueeze</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">shape</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">dim</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9unsqueezeERK5Shape10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Add a <em>unit</em> dimension to <code class="docutils literal notranslate"><span class="pre">shape</span></code> at the specified position. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><ul class="simple">
|
||
<li><p><strong>shape</strong> – The shape to unsqueeze. </p></li>
|
||
<li><p><strong>dim</strong> – The dimension where unit dimension should be added. </p></li>
|
||
</ul>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>A new shape with the added unit dimension. </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrNSt6size_tENSt6size_tE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrNSt6size_tENSt6size_tE"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrNSt6size_tENSt6size_tE"></span><span id="tensorrt_llm::runtime::ITensor::slice__SharedPtr.std::s.std::s"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1aa22bf50eff27dbfc64eed003851f4778"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9UniquePtrE" title="tensorrt_llm::runtime::ITensor::UniquePtr"><span class="n"><span class="pre">UniquePtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">slice</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tensor</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">offset</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrNSt6size_tENSt6size_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Creates a sliced view on the underlying <code class="docutils literal notranslate"><span class="pre">tensor</span></code>. The view will have the same data type as <code class="docutils literal notranslate"><span class="pre">tensor</span></code>. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><ul class="simple">
|
||
<li><p><strong>tensor</strong> – The tensor to view. </p></li>
|
||
<li><p><strong>offset</strong> – The offset of the view w.r.t. dimension 0 of the tensor. </p></li>
|
||
<li><p><strong>size</strong> – The size of the view w.r.t. dimension 0 of the tensor. </p></li>
|
||
</ul>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>A view on the <code class="docutils literal notranslate"><span class="pre">buffer</span></code>. </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrNSt6size_tENSt6size_tE">
|
||
<span id="_CPPv3I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceERR9TConstPtrNSt6size_tENSt6size_tE"></span><span id="_CPPv2I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceERR9TConstPtrNSt6size_tENSt6size_tE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TConstPtr</span></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">enable_if_t</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">is_const_v</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime18PointerElementTypeE" title="tensorrt_llm::runtime::PointerElementType"><span class="n"><span class="pre">PointerElementType</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrNSt6size_tENSt6size_tE" title="tensorrt_llm::runtime::ITensor::slice::TConstPtr"><span class="n"><span class="pre">TConstPtr</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="kt"><span class="pre">int</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a201a60f7da459ec35f5d750e8d7053c4"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor14UniqueConstPtrE" title="tensorrt_llm::runtime::ITensor::UniqueConstPtr"><span class="n"><span class="pre">UniqueConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">slice</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrNSt6size_tENSt6size_tE" title="tensorrt_llm::runtime::ITensor::slice::TConstPtr"><span class="n"><span class="pre">TConstPtr</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">tensor</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">offset</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrNSt6size_tENSt6size_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrNSt6size_tE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrNSt6size_tE"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrNSt6size_tE"></span><span id="tensorrt_llm::runtime::ITensor::slice__SharedPtr.std::s"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a3bc0148d89f892df6beff0f63ea33e6a"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9UniquePtrE" title="tensorrt_llm::runtime::ITensor::UniquePtr"><span class="n"><span class="pre">UniquePtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">slice</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tensor</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">offset</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrNSt6size_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrNSt6size_tE">
|
||
<span id="_CPPv3I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceERR9TConstPtrNSt6size_tE"></span><span id="_CPPv2I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceERR9TConstPtrNSt6size_tE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TConstPtr</span></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">enable_if_t</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">is_const_v</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime18PointerElementTypeE" title="tensorrt_llm::runtime::PointerElementType"><span class="n"><span class="pre">PointerElementType</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrNSt6size_tE" title="tensorrt_llm::runtime::ITensor::slice::TConstPtr"><span class="n"><span class="pre">TConstPtr</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="kt"><span class="pre">int</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a01d828a32b3f280c1ee64ce5211da20b"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor14UniqueConstPtrE" title="tensorrt_llm::runtime::ITensor::UniqueConstPtr"><span class="n"><span class="pre">UniqueConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">slice</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrNSt6size_tE" title="tensorrt_llm::runtime::ITensor::slice::TConstPtr"><span class="n"><span class="pre">TConstPtr</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">tensor</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">offset</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrNSt6size_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrRK5Shape9DimType64">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrRK5Shape9DimType64"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrRK5Shape9DimType64"></span><span id="tensorrt_llm::runtime::ITensor::slice__SharedPtr.ShapeCR.DimType64"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a60cd2e9b1d7c91a1201510f8deec9248"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9UniquePtrE" title="tensorrt_llm::runtime::ITensor::UniquePtr"><span class="n"><span class="pre">UniquePtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">slice</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tensor</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">offsetDims</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9DimType64E" title="tensorrt_llm::runtime::ITensor::DimType64"><span class="n"><span class="pre">DimType64</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrRK5Shape9DimType64" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><ul class="simple">
|
||
<li><p><strong>offsetDims</strong> – The offset in multiple dimensions.</p></li>
|
||
<li><p><strong>tensor</strong> – The tensor to view. </p></li>
|
||
<li><p><strong>offsetDims</strong> – The offset dimensions of the view. </p></li>
|
||
<li><p><strong>size</strong> – The size of the view w.r.t. the last dimension in offsetDims. </p></li>
|
||
<li><p><strong>offsetDims</strong> – specifies all dimensions. </p></li>
|
||
</ul>
|
||
</dd>
|
||
<dt class="field-even">Throws<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p><span><span class="cpp-expr sig sig-inline cpp"><span class="n">Whenever</span></span></span> – offset overflows or the last dimension offset+size overflows. </p>
|
||
</dd>
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>A view of shape [size, the rest dimensions] or [size] when</p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrRKNSt16initializer_listI9DimType64EE9DimType64">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrRKNSt16initializer_listI9DimType64EE9DimType64"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrRKNSt16initializer_listI9DimType64EE9DimType64"></span><span id="tensorrt_llm::runtime::ITensor::slice__SharedPtr.std::initializer_list:DimType64:CR.DimType64"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a1f486293132bbe0ac17a430db3d8c87c"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9UniquePtrE" title="tensorrt_llm::runtime::ITensor::UniquePtr"><span class="n"><span class="pre">UniquePtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">slice</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tensor</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">initializer_list</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9DimType64E" title="tensorrt_llm::runtime::ITensor::DimType64"><span class="n"><span class="pre">DimType64</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">offsetDims</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9DimType64E" title="tensorrt_llm::runtime::ITensor::DimType64"><span class="n"><span class="pre">DimType64</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrRKNSt16initializer_listI9DimType64EE9DimType64" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrRK5ShapeNSt6size_tE">
|
||
<span id="_CPPv3I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceERR9TConstPtrRK5ShapeNSt6size_tE"></span><span id="_CPPv2I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceERR9TConstPtrRK5ShapeNSt6size_tE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TConstPtr</span></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">enable_if_t</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">is_const_v</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime18PointerElementTypeE" title="tensorrt_llm::runtime::PointerElementType"><span class="n"><span class="pre">PointerElementType</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrRK5ShapeNSt6size_tE" title="tensorrt_llm::runtime::ITensor::slice::TConstPtr"><span class="n"><span class="pre">TConstPtr</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="kt"><span class="pre">int</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a2d0ab0192c27695bedeb7134cad400ac"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor14UniqueConstPtrE" title="tensorrt_llm::runtime::ITensor::UniqueConstPtr"><span class="n"><span class="pre">UniqueConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">slice</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrRK5ShapeNSt6size_tE" title="tensorrt_llm::runtime::ITensor::slice::TConstPtr"><span class="n"><span class="pre">TConstPtr</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">tensor</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">offsetDims</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrRK5ShapeNSt6size_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrRKNSt16initializer_listI9DimType64EENSt6size_tE">
|
||
<span id="_CPPv3I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceERR9TConstPtrRKNSt16initializer_listI9DimType64EENSt6size_tE"></span><span id="_CPPv2I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceERR9TConstPtrRKNSt16initializer_listI9DimType64EENSt6size_tE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TConstPtr</span></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">enable_if_t</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">is_const_v</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime18PointerElementTypeE" title="tensorrt_llm::runtime::PointerElementType"><span class="n"><span class="pre">PointerElementType</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrRKNSt16initializer_listI9DimType64EENSt6size_tE" title="tensorrt_llm::runtime::ITensor::slice::TConstPtr"><span class="n"><span class="pre">TConstPtr</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="kt"><span class="pre">int</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1ad35b215bf9822c1e21c77647c78ba4ec"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor14UniqueConstPtrE" title="tensorrt_llm::runtime::ITensor::UniqueConstPtr"><span class="n"><span class="pre">UniqueConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">slice</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrRKNSt16initializer_listI9DimType64EENSt6size_tE" title="tensorrt_llm::runtime::ITensor::slice::TConstPtr"><span class="n"><span class="pre">TConstPtr</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">tensor</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">initializer_list</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9DimType64E" title="tensorrt_llm::runtime::ITensor::DimType64"><span class="n"><span class="pre">DimType64</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">offsetDims</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrRKNSt16initializer_listI9DimType64EENSt6size_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrRK5Shape">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrRK5Shape"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrRK5Shape"></span><span id="tensorrt_llm::runtime::ITensor::slice__SharedPtr.ShapeCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1ade7e339e615cd675e4201903b3d28139"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9UniquePtrE" title="tensorrt_llm::runtime::ITensor::UniquePtr"><span class="n"><span class="pre">UniquePtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">slice</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tensor</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">offsetDims</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrRK5Shape" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>return the rest slices at the last dimension when <code class="docutils literal notranslate"><span class="pre">size</span></code> omitted. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrRKNSt16initializer_listI9DimType64EE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrRKNSt16initializer_listI9DimType64EE"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrRKNSt16initializer_listI9DimType64EE"></span><span id="tensorrt_llm::runtime::ITensor::slice__SharedPtr.std::initializer_list:DimType64:CR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a612d0d2c27d219f5344a5acb244faca3"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9UniquePtrE" title="tensorrt_llm::runtime::ITensor::UniquePtr"><span class="n"><span class="pre">UniquePtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">slice</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tensor</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">initializer_list</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9DimType64E" title="tensorrt_llm::runtime::ITensor::DimType64"><span class="n"><span class="pre">DimType64</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">offsetDims</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5sliceE9SharedPtrRKNSt16initializer_listI9DimType64EE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrRK5Shape">
|
||
<span id="_CPPv3I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceERR9TConstPtrRK5Shape"></span><span id="_CPPv2I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceERR9TConstPtrRK5Shape"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TConstPtr</span></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">enable_if_t</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">is_const_v</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime18PointerElementTypeE" title="tensorrt_llm::runtime::PointerElementType"><span class="n"><span class="pre">PointerElementType</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrRK5Shape" title="tensorrt_llm::runtime::ITensor::slice::TConstPtr"><span class="n"><span class="pre">TConstPtr</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="kt"><span class="pre">int</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1aaad4e79382fcd3fc2f55c0f4d088dc74"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor14UniqueConstPtrE" title="tensorrt_llm::runtime::ITensor::UniqueConstPtr"><span class="n"><span class="pre">UniqueConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">slice</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrRK5Shape" title="tensorrt_llm::runtime::ITensor::slice::TConstPtr"><span class="n"><span class="pre">TConstPtr</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">tensor</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">offsetDims</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrRK5Shape" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrRKNSt16initializer_listI9DimType64EE">
|
||
<span id="_CPPv3I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceERR9TConstPtrRKNSt16initializer_listI9DimType64EE"></span><span id="_CPPv2I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceERR9TConstPtrRKNSt16initializer_listI9DimType64EE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TConstPtr</span></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">enable_if_t</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">is_const_v</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime18PointerElementTypeE" title="tensorrt_llm::runtime::PointerElementType"><span class="n"><span class="pre">PointerElementType</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrRKNSt16initializer_listI9DimType64EE" title="tensorrt_llm::runtime::ITensor::slice::TConstPtr"><span class="n"><span class="pre">TConstPtr</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="kt"><span class="pre">int</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1aa61976020990ef44546e22921156dd88"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor14UniqueConstPtrE" title="tensorrt_llm::runtime::ITensor::UniqueConstPtr"><span class="n"><span class="pre">UniqueConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">slice</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrRKNSt16initializer_listI9DimType64EE" title="tensorrt_llm::runtime::ITensor::slice::TConstPtr"><span class="n"><span class="pre">TConstPtr</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">tensor</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">initializer_list</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9DimType64E" title="tensorrt_llm::runtime::ITensor::DimType64"><span class="n"><span class="pre">DimType64</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">offsetDims</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor5sliceE14UniqueConstPtrRR9TConstPtrRKNSt16initializer_listI9DimType64EE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor2atE9SharedPtrRK5Shape">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor2atE9SharedPtrRK5Shape"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor2atE9SharedPtrRK5Shape"></span><span id="tensorrt_llm::runtime::ITensor::at__SharedPtr.ShapeCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a02711c850a3a20114316a76ac00593a1"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9UniquePtrE" title="tensorrt_llm::runtime::ITensor::UniquePtr"><span class="n"><span class="pre">UniquePtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">at</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tensor</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">offsetDims</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor2atE9SharedPtrRK5Shape" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>offsetDims</strong> – specifies all dimensions. </p>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>Just the block at the point, with shape of [the rest dimensions] or [1] when</p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor2atE9SharedPtrRKNSt16initializer_listI9DimType64EE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor2atE9SharedPtrRKNSt16initializer_listI9DimType64EE"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor2atE9SharedPtrRKNSt16initializer_listI9DimType64EE"></span><span id="tensorrt_llm::runtime::ITensor::at__SharedPtr.std::initializer_list:DimType64:CR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a920d45c26d8b49480daf695779c67183"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9UniquePtrE" title="tensorrt_llm::runtime::ITensor::UniquePtr"><span class="n"><span class="pre">UniquePtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">at</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tensor</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">initializer_list</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9DimType64E" title="tensorrt_llm::runtime::ITensor::DimType64"><span class="n"><span class="pre">DimType64</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">offsetDims</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor2atE9SharedPtrRKNSt16initializer_listI9DimType64EE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor2atE14UniqueConstPtrRR9TConstPtrRK5Shape">
|
||
<span id="_CPPv3I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor2atERR9TConstPtrRK5Shape"></span><span id="_CPPv2I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor2atERR9TConstPtrRK5Shape"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TConstPtr</span></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">enable_if_t</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">is_const_v</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime18PointerElementTypeE" title="tensorrt_llm::runtime::PointerElementType"><span class="n"><span class="pre">PointerElementType</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor2atE14UniqueConstPtrRR9TConstPtrRK5Shape" title="tensorrt_llm::runtime::ITensor::at::TConstPtr"><span class="n"><span class="pre">TConstPtr</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="kt"><span class="pre">int</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1ab0e745de500a91974a20dd31783f3efb"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor14UniqueConstPtrE" title="tensorrt_llm::runtime::ITensor::UniqueConstPtr"><span class="n"><span class="pre">UniqueConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">at</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor2atE14UniqueConstPtrRR9TConstPtrRK5Shape" title="tensorrt_llm::runtime::ITensor::at::TConstPtr"><span class="n"><span class="pre">TConstPtr</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">tensor</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">offsetDims</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor2atE14UniqueConstPtrRR9TConstPtrRK5Shape" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor2atEN7ITensor14UniqueConstPtrERR9TConstPtrRKNSt16initializer_listI9DimType64EE">
|
||
<span id="_CPPv3I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor2atERR9TConstPtrRKNSt16initializer_listI9DimType64EE"></span><span id="_CPPv2I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor2atERR9TConstPtrRKNSt16initializer_listI9DimType64EE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TConstPtr</span></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">enable_if_t</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">is_const_v</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime18PointerElementTypeE" title="tensorrt_llm::runtime::PointerElementType"><span class="n"><span class="pre">PointerElementType</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor2atEN7ITensor14UniqueConstPtrERR9TConstPtrRKNSt16initializer_listI9DimType64EE" title="tensorrt_llm::runtime::ITensor::at::TConstPtr"><span class="n"><span class="pre">TConstPtr</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="kt"><span class="pre">int</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a60ebb8d28d2e24cf31f941645b237552"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor14UniqueConstPtrE" title="tensorrt_llm::runtime::ITensor::UniqueConstPtr"><span class="n"><span class="pre">UniqueConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">at</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor2atEN7ITensor14UniqueConstPtrERR9TConstPtrRKNSt16initializer_listI9DimType64EE" title="tensorrt_llm::runtime::ITensor::at::TConstPtr"><span class="n"><span class="pre">TConstPtr</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">tensor</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">initializer_list</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9DimType64E" title="tensorrt_llm::runtime::ITensor::DimType64"><span class="n"><span class="pre">DimType64</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">offsetDims</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor2atEN7ITensor14UniqueConstPtrERR9TConstPtrRKNSt16initializer_listI9DimType64EE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor4viewEN7IBuffer9SharedPtrERK5Shape">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor4viewEN7IBuffer9SharedPtrERK5Shape"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor4viewEN7IBuffer9SharedPtrERK5Shape"></span><span id="tensorrt_llm::runtime::ITensor::view__IBuffer::SharedPtr.ShapeCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1aeb7a31c92f9ebc17e4648ba4fff7e16e"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9UniquePtrE" title="tensorrt_llm::runtime::ITensor::UniquePtr"><span class="n"><span class="pre">UniquePtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">view</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer9SharedPtrE" title="tensorrt_llm::runtime::IBuffer::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">buffer</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">dims</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor4viewEN7IBuffer9SharedPtrERK5Shape" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Returns a view on the underlying <code class="docutils literal notranslate"><span class="pre">buffer</span></code> (or tensor) with the given shape. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><ul class="simple">
|
||
<li><p><strong>tensor</strong> – The tensor to view. </p></li>
|
||
<li><p><strong>shape</strong> – The shape of the view. </p></li>
|
||
</ul>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>A view on the <code class="docutils literal notranslate"><span class="pre">tensor</span></code>. </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor4viewE14UniqueConstPtrRR9TConstPtrRK5Shape">
|
||
<span id="_CPPv3I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor4viewERR9TConstPtrRK5Shape"></span><span id="_CPPv2I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor4viewERR9TConstPtrRK5Shape"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TConstPtr</span></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">enable_if_t</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">is_const_v</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime18PointerElementTypeE" title="tensorrt_llm::runtime::PointerElementType"><span class="n"><span class="pre">PointerElementType</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor4viewE14UniqueConstPtrRR9TConstPtrRK5Shape" title="tensorrt_llm::runtime::ITensor::view::TConstPtr"><span class="n"><span class="pre">TConstPtr</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="kt"><span class="pre">int</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a481831d7c0d45fe462e91b07beef2617"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor14UniqueConstPtrE" title="tensorrt_llm::runtime::ITensor::UniqueConstPtr"><span class="n"><span class="pre">UniqueConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">view</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor4viewE14UniqueConstPtrRR9TConstPtrRK5Shape" title="tensorrt_llm::runtime::ITensor::view::TConstPtr"><span class="n"><span class="pre">TConstPtr</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">tensor</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">dims</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7ITensor4viewE14UniqueConstPtrRR9TConstPtrRK5Shape" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor4viewE9SharedPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor4viewE9SharedPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor4viewE9SharedPtr"></span><span id="tensorrt_llm::runtime::ITensor::view__SharedPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a0d748a6fdb8219af29ba918878bf44b2"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9UniquePtrE" title="tensorrt_llm::runtime::ITensor::UniquePtr"><span class="n"><span class="pre">UniquePtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">view</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tensor</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor4viewE9SharedPtr" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Returns a view on the underlying <code class="docutils literal notranslate"><span class="pre">tensor</span></code> which can be independently reshaped. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>tensor</strong> – The tensor to view. </p>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>A view on the <code class="docutils literal notranslate"><span class="pre">tensor</span></code>. </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor8flattenNE9SharedPtrNSt7int64_tE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor8flattenNE9SharedPtrNSt7int64_tE"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor8flattenNE9SharedPtrNSt7int64_tE"></span><span id="tensorrt_llm::runtime::ITensor::flattenN__SharedPtr.std::int64_t"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a7078d3b5b40ba1257c3be030e4766c9e"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9UniquePtrE" title="tensorrt_llm::runtime::ITensor::UniquePtr"><span class="n"><span class="pre">UniquePtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">flattenN</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tensor</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">int64_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">sliceN</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="o"><span class="pre">-</span></span><span class="m"><span class="pre">1</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor8flattenNE9SharedPtrNSt7int64_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Returns a flattened view on the underlying <code class="docutils literal notranslate"><span class="pre">tensor</span></code> which can be independently reshaped. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><ul class="simple">
|
||
<li><p><strong>tensor</strong> – The tensor to flatten. </p></li>
|
||
<li><p><strong>sliceN</strong> – Slice the first N elements after flattening. -1 means take the whole flattened tensor. </p></li>
|
||
</ul>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>A flatten view on the <code class="docutils literal notranslate"><span class="pre">tensor</span></code>. </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor4wrapEPvN8nvinfer18DataTypeERK5ShapeNSt6size_tE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor4wrapEPvN8nvinfer18DataTypeERK5ShapeNSt6size_tE"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor4wrapEPvN8nvinfer18DataTypeERK5ShapeNSt6size_tE"></span><span id="tensorrt_llm::runtime::ITensor::wrap__voidP.nvinfer1::DataType.ShapeCR.std::s"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1abb208b99fc6a404a13215fe2d43637a4"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9UniquePtrE" title="tensorrt_llm::runtime::ITensor::UniquePtr"><span class="n"><span class="pre">UniquePtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">wrap</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="n sig-param"><span class="pre">data</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">type</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">shape</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">capacity</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor4wrapEPvN8nvinfer18DataTypeERK5ShapeNSt6size_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Wraps the given <code class="docutils literal notranslate"><span class="pre">data</span></code> in an <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1ITensor"><span class="std std-ref"><span class="pre">ITensor</span></span></a></code>. The <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1ITensor"><span class="std std-ref"><span class="pre">ITensor</span></span></a></code> will not own the underlying <code class="docutils literal notranslate"><span class="pre">data</span></code> and cannot be reshaped beyond <code class="docutils literal notranslate"><span class="pre">capacity</span></code>. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><ul class="simple">
|
||
<li><p><strong>data</strong> – The data to wrap. </p></li>
|
||
<li><p><strong>type</strong> – The data type of the <code class="docutils literal notranslate"><span class="pre">data</span></code>. </p></li>
|
||
<li><p><strong>shape</strong> – The shape of the tensor. </p></li>
|
||
<li><p><strong>capacity</strong> – The capacity of the buffer. </p></li>
|
||
</ul>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>An <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1ITensor"><span class="std std-ref"><span class="pre">ITensor</span></span></a></code>. </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor4wrapEPvN8nvinfer18DataTypeERK5Shape">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor4wrapEPvN8nvinfer18DataTypeERK5Shape"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor4wrapEPvN8nvinfer18DataTypeERK5Shape"></span><span id="tensorrt_llm::runtime::ITensor::wrap__voidP.nvinfer1::DataType.ShapeCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a5fdde0b2cc5bb0ee07c042964bb48c85"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9UniquePtrE" title="tensorrt_llm::runtime::ITensor::UniquePtr"><span class="n"><span class="pre">UniquePtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">wrap</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="n sig-param"><span class="pre">data</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">type</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">shape</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor4wrapEPvN8nvinfer18DataTypeERK5Shape" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime7ITensor4wrapE9UniquePtrP1TRK5ShapeNSt6size_tE">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime7ITensor4wrapEP1TRK5ShapeNSt6size_tE"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime7ITensor4wrapEP1TRK5ShapeNSt6size_tE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a6cb1dc1fe13779d7d0bed2062df2c0e0"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9UniquePtrE" title="tensorrt_llm::runtime::ITensor::UniquePtr"><span class="n"><span class="pre">UniquePtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">wrap</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime7ITensor4wrapE9UniquePtrP1TRK5ShapeNSt6size_tE" title="tensorrt_llm::runtime::ITensor::wrap::T"><span class="n"><span class="pre">T</span></span></a><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="n sig-param"><span class="pre">data</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">shape</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">capacity</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime7ITensor4wrapE9UniquePtrP1TRK5ShapeNSt6size_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime7ITensor4wrapE9UniquePtrP1TRK5Shape">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime7ITensor4wrapEP1TRK5Shape"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime7ITensor4wrapEP1TRK5Shape"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a496e23f8dcc29f792a7e0f70fca11973"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9UniquePtrE" title="tensorrt_llm::runtime::ITensor::UniquePtr"><span class="n"><span class="pre">UniquePtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">wrap</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime7ITensor4wrapE9UniquePtrP1TRK5Shape" title="tensorrt_llm::runtime::ITensor::wrap::T"><span class="n"><span class="pre">T</span></span></a><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="n sig-param"><span class="pre">data</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">shape</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime7ITensor4wrapE9UniquePtrP1TRK5Shape" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime7ITensor4wrapE9UniquePtrRNSt6vectorI1TEERK5Shape">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime7ITensor4wrapERNSt6vectorI1TEERK5Shape"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime7ITensor4wrapERNSt6vectorI1TEERK5Shape"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1ab27a7a0ae81205699d2b2bc618367857"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9UniquePtrE" title="tensorrt_llm::runtime::ITensor::UniquePtr"><span class="n"><span class="pre">UniquePtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">wrap</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime7ITensor4wrapE9UniquePtrRNSt6vectorI1TEERK5Shape" title="tensorrt_llm::runtime::ITensor::wrap::T"><span class="n"><span class="pre">T</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">v</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">shape</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime7ITensor4wrapE9UniquePtrRNSt6vectorI1TEERK5Shape" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor9makeShapeERKNSt16initializer_listI9DimType64EE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor9makeShapeERKNSt16initializer_listI9DimType64EE"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor9makeShapeERKNSt16initializer_listI9DimType64EE"></span><span id="tensorrt_llm::runtime::ITensor::makeShape__std::initializer_list:DimType64:CR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a6ae0b710b39f4f5d821681188c8afce4"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">makeShape</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">initializer_list</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9DimType64E" title="tensorrt_llm::runtime::ITensor::DimType64"><span class="n"><span class="pre">DimType64</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">dims</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9makeShapeERKNSt16initializer_listI9DimType64EE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>A convenience function to create a tensor shape with the given dimensions. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor8toStringERK5Shape">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor8toStringERK5Shape"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor8toStringERK5Shape"></span><span id="tensorrt_llm::runtime::ITensor::toString__ShapeCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1af9ba06b8cbf144ea573e6767af85bd43"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">toString</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">dims</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor8toStringERK5Shape" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>A convenience function for converting a tensor shape to a <code class="docutils literal notranslate"><span class="pre">string</span></code>. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor11shapeEqualsERK5ShapeRK5Shape">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor11shapeEqualsERK5ShapeRK5Shape"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor11shapeEqualsERK5ShapeRK5Shape"></span><span id="tensorrt_llm::runtime::ITensor::shapeEquals__ShapeCR.ShapeCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1ae39159cc236f70454983d0193570e1c4"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">shapeEquals</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">lhs</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">rhs</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor11shapeEqualsERK5ShapeRK5Shape" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>A convenience function to compare shapes. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime7ITensor11shapeEqualsEbRK5ShapePK1T10SizeType32">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime7ITensor11shapeEqualsERK5ShapePK1T10SizeType32"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime7ITensor11shapeEqualsERK5ShapePK1T10SizeType32"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1afbf1a7c0ee57cf176f6dfbb04766b9d9"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">shapeEquals</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor5ShapeE" title="tensorrt_llm::runtime::ITensor::Shape"><span class="n"><span class="pre">Shape</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">lhs</span></span>, <a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime7ITensor11shapeEqualsEbRK5ShapePK1T10SizeType32" title="tensorrt_llm::runtime::ITensor::shapeEquals::T"><span class="n"><span class="pre">T</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="n sig-param"><span class="pre">dims</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">count</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime7ITensor11shapeEqualsEbRK5ShapePK1T10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>A convenience function to compare shapes. </p>
|
||
</dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-protected-functions">Protected Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor7ITensorEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor7ITensorEv"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor7ITensorEv"></span><span id="tensorrt_llm::runtime::ITensor::ITensor"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a60742a4752e9d9d5a82fff9e24518953"></span><span class="sig-name descname"><span class="n"><span class="pre">ITensor</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">default</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor7ITensorEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-protected-static-functions">Protected Static Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor8castSizeE6size_t">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor8castSizeE6size_t"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor8castSizeE6size_t"></span><span id="tensorrt_llm::runtime::ITensor::castSize__s"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a0d4cb08fc550e2b0e97399a76e3d64ec"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9DimType64E" title="tensorrt_llm::runtime::ITensor::DimType64"><span class="n"><span class="pre">DimType64</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">castSize</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">newSize</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7ITensor8castSizeE6size_t" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-friends">Friends</p>
|
||
<dl>
|
||
<dt class="sig sig-object cpp">
|
||
<em class="property"><span class="pre">friend</span> <span class="pre">class</span></em> <span class="pre">ITensorBindings</span></dt>
|
||
</dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="ipcutils-h">
|
||
<h2>ipcUtils.h<a class="headerlink" href="#ipcutils-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-functions">Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime20lamportInitializeAllEPvPvPv6size_t">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime20lamportInitializeAllEPvPvPv6size_t"></span><span id="_CPPv2N12tensorrt_llm7runtime20lamportInitializeAllEPvPvPv6size_t"></span><span id="tensorrt_llm::runtime::lamportInitializeAll__voidP.voidP.voidP.s"></span><span class="target" id="ipcUtils_8h_1a40562e2c0ec119fa1918eb42cef0b074"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">lamportInitializeAll</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="n sig-param"><span class="pre">buffer_0</span></span>, <span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="n sig-param"><span class="pre">buffer_1</span></span>, <span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="n sig-param"><span class="pre">buffer_2</span></span>, <span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime20lamportInitializeAllEPvPvPv6size_t" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime16AllReduceBuffersE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16AllReduceBuffersE"></span><span id="_CPPv2N12tensorrt_llm7runtime16AllReduceBuffersE"></span><span id="tensorrt_llm::runtime::AllReduceBuffers"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1AllReduceBuffers"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">AllReduceBuffers</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime16AllReduceBuffersE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime16AllReduceBuffers9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16AllReduceBuffers9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime16AllReduceBuffers9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1AllReduceBuffers_1a785db934d1fa831386655fff122f594e"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime16AllReduceBuffers9TensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime16AllReduceBuffers16AllReduceBuffersE10SizeType3210SizeType3210SizeType3210SizeType32RK13BufferManagerRK11WorldConfigKb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16AllReduceBuffers16AllReduceBuffersE10SizeType3210SizeType3210SizeType3210SizeType32RK13BufferManagerRK11WorldConfigKb"></span><span id="_CPPv2N12tensorrt_llm7runtime16AllReduceBuffers16AllReduceBuffersE10SizeType3210SizeType3210SizeType3210SizeType32RK13BufferManagerRK11WorldConfigKb"></span><span id="tensorrt_llm::runtime::AllReduceBuffers::AllReduceBuffers__SizeType32.SizeType32.SizeType32.SizeType32.BufferManagerCR.WorldConfigCR.bC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1AllReduceBuffers_1ab531d14cdf9443f3730aa2aed8e1635e"></span><span class="sig-name descname"><span class="n"><span class="pre">AllReduceBuffers</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxBatchSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxBeamWidth</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxSequenceLength</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">hiddenSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManagerE" title="tensorrt_llm::runtime::BufferManager"><span class="n"><span class="pre">BufferManager</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">manager</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfigE" title="tensorrt_llm::runtime::WorldConfig"><span class="n"><span class="pre">WorldConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">worldConfig</span></span>, <span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">fakeBuffers</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">false</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime16AllReduceBuffers16AllReduceBuffersE10SizeType3210SizeType3210SizeType3210SizeType32RK13BufferManagerRK11WorldConfigKb" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime16AllReduceBuffers18mAllReduceCommPtrsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16AllReduceBuffers18mAllReduceCommPtrsE"></span><span id="_CPPv2N12tensorrt_llm7runtime16AllReduceBuffers18mAllReduceCommPtrsE"></span><span id="tensorrt_llm::runtime::AllReduceBuffers::mAllReduceCommPtrs__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1AllReduceBuffers_1ab48e63279d11f42d71c3621820d2520c"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16AllReduceBuffers9TensorPtrE" title="tensorrt_llm::runtime::AllReduceBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mAllReduceCommPtrs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime16AllReduceBuffers18mAllReduceCommPtrsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime16AllReduceBuffers17mIpcMemoryHandlesE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16AllReduceBuffers17mIpcMemoryHandlesE"></span><span id="_CPPv2N12tensorrt_llm7runtime16AllReduceBuffers17mIpcMemoryHandlesE"></span><span id="tensorrt_llm::runtime::AllReduceBuffers::mIpcMemoryHandles__std::vector:runtime::IpcMemory:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1AllReduceBuffers_1a162c983f7dc981a8c4af57510637e767"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemoryE" title="tensorrt_llm::runtime::IpcMemory"><span class="n"><span class="pre">IpcMemory</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mIpcMemoryHandles</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime16AllReduceBuffers17mIpcMemoryHandlesE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9IpcMemoryE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9IpcMemoryE"></span><span id="_CPPv2N12tensorrt_llm7runtime9IpcMemoryE"></span><span id="tensorrt_llm::runtime::IpcMemory"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IpcMemory"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">IpcMemory</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemoryE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9IpcMemory9BufferPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9IpcMemory9BufferPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime9IpcMemory9BufferPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IpcMemory_1ab1b59b5e9ca9bae538f4f96f67f54b4c"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">BufferPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer9SharedPtrE" title="tensorrt_llm::runtime::IBuffer::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory9BufferPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9IpcMemory9IpcMemoryENSt6size_tERK13BufferManagerRK11WorldConfigb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9IpcMemory9IpcMemoryENSt6size_tERK13BufferManagerRK11WorldConfigb"></span><span id="_CPPv2N12tensorrt_llm7runtime9IpcMemory9IpcMemoryENSt6size_tERK13BufferManagerRK11WorldConfigb"></span><span id="tensorrt_llm::runtime::IpcMemory::IpcMemory__std::s.BufferManagerCR.WorldConfigCR.b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IpcMemory_1a3375828b692be378adbab4475b734f54"></span><span class="sig-name descname"><span class="n"><span class="pre">IpcMemory</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">bufferSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManagerE" title="tensorrt_llm::runtime::BufferManager"><span class="n"><span class="pre">BufferManager</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">manager</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfigE" title="tensorrt_llm::runtime::WorldConfig"><span class="n"><span class="pre">WorldConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">worldConfig</span></span>, <span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">openIpc</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">true</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory9IpcMemoryENSt6size_tERK13BufferManagerRK11WorldConfigb" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9IpcMemoryD0Ev">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9IpcMemoryD0Ev"></span><span id="_CPPv2N12tensorrt_llm7runtime9IpcMemoryD0Ev"></span><span id="tensorrt_llm::runtime::IpcMemory::~IpcMemory"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IpcMemory_1a0f068272bfaeadb2e976b44adf47c484"></span><span class="sig-name descname"><span class="n"><span class="pre">~IpcMemory</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemoryD0Ev" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9IpcMemory9IpcMemoryERK9IpcMemory">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9IpcMemory9IpcMemoryERK9IpcMemory"></span><span id="_CPPv2N12tensorrt_llm7runtime9IpcMemory9IpcMemoryERK9IpcMemory"></span><span id="tensorrt_llm::runtime::IpcMemory::IpcMemory__IpcMemoryCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IpcMemory_1ae3a748873dec82811c4b0014df78d107"></span><span class="sig-name descname"><span class="n"><span class="pre">IpcMemory</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory9IpcMemoryERK9IpcMemory" title="tensorrt_llm::runtime::IpcMemory::IpcMemory"><span class="n"><span class="pre">IpcMemory</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="p"><span class="pre">&</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">delete</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory9IpcMemoryERK9IpcMemory" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9IpcMemoryaSERK9IpcMemory">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9IpcMemoryaSERK9IpcMemory"></span><span id="_CPPv2N12tensorrt_llm7runtime9IpcMemoryaSERK9IpcMemory"></span><span id="tensorrt_llm::runtime::IpcMemory::assign-operator__IpcMemoryCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IpcMemory_1ad9a357f98e8909799d922dce5f777bd7"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemoryE" title="tensorrt_llm::runtime::IpcMemory"><span class="n"><span class="pre">IpcMemory</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="k"><span class="pre">operator</span></span><span class="o"><span class="pre">=</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemoryE" title="tensorrt_llm::runtime::IpcMemory"><span class="n"><span class="pre">IpcMemory</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="p"><span class="pre">&</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">delete</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemoryaSERK9IpcMemory" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9IpcMemory9IpcMemoryERR9IpcMemory">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9IpcMemory9IpcMemoryERR9IpcMemory"></span><span id="_CPPv2N12tensorrt_llm7runtime9IpcMemory9IpcMemoryERR9IpcMemory"></span><span id="tensorrt_llm::runtime::IpcMemory::IpcMemory__IpcMemoryRR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IpcMemory_1a611381953a05d2e92e68831e5a459b1a"></span><span class="sig-name descname"><span class="n"><span class="pre">IpcMemory</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory9IpcMemoryERR9IpcMemory" title="tensorrt_llm::runtime::IpcMemory::IpcMemory"><span class="n"><span class="pre">IpcMemory</span></span></a><span class="p"><span class="pre">&</span></span><span class="p"><span class="pre">&</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">default</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory9IpcMemoryERR9IpcMemory" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9IpcMemoryaSERR9IpcMemory">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9IpcMemoryaSERR9IpcMemory"></span><span id="_CPPv2N12tensorrt_llm7runtime9IpcMemoryaSERR9IpcMemory"></span><span id="tensorrt_llm::runtime::IpcMemory::assign-operator__IpcMemoryRR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IpcMemory_1acd3552f67bf6779ee90c0b2dd471cca4"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemoryE" title="tensorrt_llm::runtime::IpcMemory"><span class="n"><span class="pre">IpcMemory</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="k"><span class="pre">operator</span></span><span class="o"><span class="pre">=</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemoryE" title="tensorrt_llm::runtime::IpcMemory"><span class="n"><span class="pre">IpcMemory</span></span></a><span class="p"><span class="pre">&</span></span><span class="p"><span class="pre">&</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">default</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemoryaSERR9IpcMemory" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime9IpcMemory11getCommPtrsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime9IpcMemory11getCommPtrsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime9IpcMemory11getCommPtrsEv"></span><span id="tensorrt_llm::runtime::IpcMemory::getCommPtrsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IpcMemory_1ad92163a8d4f3b895654e67222be18aac"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="kt"><span class="pre">void</span></span><span class="p"><span class="pre">*</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="n"><span class="pre">getCommPtrs</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime9IpcMemory11getCommPtrsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9IpcMemory10FLAGS_SIZEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9IpcMemory10FLAGS_SIZEE"></span><span id="_CPPv2N12tensorrt_llm7runtime9IpcMemory10FLAGS_SIZEE"></span><span id="tensorrt_llm::runtime::IpcMemory::FLAGS_SIZE__s"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IpcMemory_1a3770f9deca8b19095cbe4f0268664265"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">FLAGS_SIZE</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">(</span></span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">kernels</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">MAX_ALL_REDUCE_BLOCKS</span></span><span class="w"> </span><span class="o"><span class="pre">+</span></span><span class="w"> </span><span class="m"><span class="pre">1</span></span><span class="p"><span class="pre">)</span></span><span class="w"> </span><span class="o"><span class="pre">*</span></span><span class="w"> </span><span class="k"><span class="pre">sizeof</span></span><span class="p"><span class="pre">(</span></span><span class="n"><span class="pre">uint32_t</span></span><span class="p"><span class="pre">)</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory10FLAGS_SIZEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-functions">Private Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9IpcMemory17allocateIpcMemoryENSt6size_tERK13BufferManagerRK11WorldConfig">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9IpcMemory17allocateIpcMemoryENSt6size_tERK13BufferManagerRK11WorldConfig"></span><span id="_CPPv2N12tensorrt_llm7runtime9IpcMemory17allocateIpcMemoryENSt6size_tERK13BufferManagerRK11WorldConfig"></span><span id="tensorrt_llm::runtime::IpcMemory::allocateIpcMemory__std::s.BufferManagerCR.WorldConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IpcMemory_1a68894ef110ccba3296dde9e7dff72e61"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">allocateIpcMemory</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">bufferSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManagerE" title="tensorrt_llm::runtime::BufferManager"><span class="n"><span class="pre">BufferManager</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">manager</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfigE" title="tensorrt_llm::runtime::WorldConfig"><span class="n"><span class="pre">WorldConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">worldConfig</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory17allocateIpcMemoryENSt6size_tERK13BufferManagerRK11WorldConfig" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9IpcMemory16destroyIpcMemoryEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9IpcMemory16destroyIpcMemoryEv"></span><span id="_CPPv2N12tensorrt_llm7runtime9IpcMemory16destroyIpcMemoryEv"></span><span id="tensorrt_llm::runtime::IpcMemory::destroyIpcMemory"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IpcMemory_1a7b7c035819f96cb37702472ea179c33b"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">destroyIpcMemory</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory16destroyIpcMemoryEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-members">Private Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9IpcMemory7mTpRankE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9IpcMemory7mTpRankE"></span><span id="_CPPv2N12tensorrt_llm7runtime9IpcMemory7mTpRankE"></span><span id="tensorrt_llm::runtime::IpcMemory::mTpRank__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IpcMemory_1a9ce83e03d280dad785bc830c997dad90"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mTpRank</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory7mTpRankE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9IpcMemory9mCommPtrsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9IpcMemory9mCommPtrsE"></span><span id="_CPPv2N12tensorrt_llm7runtime9IpcMemory9mCommPtrsE"></span><span id="tensorrt_llm::runtime::IpcMemory::mCommPtrs__std::vector:voidP:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IpcMemory_1aa15ac983a61b79976435b4d067daccc2"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="kt"><span class="pre">void</span></span><span class="p"><span class="pre">*</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mCommPtrs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory9mCommPtrsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9IpcMemory7mBufferE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9IpcMemory7mBufferE"></span><span id="_CPPv2N12tensorrt_llm7runtime9IpcMemory7mBufferE"></span><span id="tensorrt_llm::runtime::IpcMemory::mBuffer__BufferPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IpcMemory_1a7907761a17efa256fdbcbe8bb95a26b2"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory9BufferPtrE" title="tensorrt_llm::runtime::IpcMemory::BufferPtr"><span class="n"><span class="pre">BufferPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mBuffer</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory7mBufferE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9IpcMemory8mOpenIpcE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9IpcMemory8mOpenIpcE"></span><span id="_CPPv2N12tensorrt_llm7runtime9IpcMemory8mOpenIpcE"></span><span id="tensorrt_llm::runtime::IpcMemory::mOpenIpc__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IpcMemory_1a3486fbaa3a609c6e364f5e29557ec630"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mOpenIpc</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory8mOpenIpcE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="loracache-h">
|
||
<h2>loraCache.h<a class="headerlink" href="#loracache-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-functions">Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9to_stringERKN9LoraCache21TaskLayerModuleConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9to_stringERKN9LoraCache21TaskLayerModuleConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime9to_stringERKN9LoraCache21TaskLayerModuleConfigE"></span><span id="tensorrt_llm::runtime::to_string__LoraCache::TaskLayerModuleConfigCR"></span><span class="target" id="loraCache_8h_1ae685625093bdac0fb98a2ef13aa1a81e"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">to_string</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCacheE" title="tensorrt_llm::runtime::LoraCache"><span class="n"><span class="pre">LoraCache</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfigE" title="tensorrt_llm::runtime::LoraCache::TaskLayerModuleConfig"><span class="n"><span class="pre">TaskLayerModuleConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">v</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9to_stringERKN9LoraCache21TaskLayerModuleConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtimelsERNSt7ostreamERKN9LoraCache21TaskLayerModuleConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtimelsERNSt7ostreamERKN9LoraCache21TaskLayerModuleConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtimelsERNSt7ostreamERKN9LoraCache21TaskLayerModuleConfigE"></span><span id="tensorrt_llm::runtime::lshift-operator__osR.LoraCache::TaskLayerModuleConfigCR"></span><span class="target" id="loraCache_8h_1a1455e26c47458506ad54c1140ac0776c"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ostream</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="k"><span class="pre">operator</span></span><span class="o"><span class="pre"><<</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ostream</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">os</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCacheE" title="tensorrt_llm::runtime::LoraCache"><span class="n"><span class="pre">LoraCache</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfigE" title="tensorrt_llm::runtime::LoraCache::TaskLayerModuleConfig"><span class="n"><span class="pre">TaskLayerModuleConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">v</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtimelsERNSt7ostreamERKN9LoraCache21TaskLayerModuleConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCacheE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCacheE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCacheE"></span><span id="tensorrt_llm::runtime::LoraCache"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">LoraCache</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCacheE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="docutils container">
|
||
<em>#include <loraCache.h></em></div>
|
||
<p><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1LoraCache"><span class="std std-ref">LoraCache</span></a></p>
|
||
<p>Caches LoRA weights with LRU eviction policy.</p>
|
||
<p>Tasks put in the cache are marked in progress and can not be evicted, until they are marked done.</p>
|
||
<p>A cache page holds a optimally sized LoRA. A page is of size [numSlots x pageWidth] An optimally size LoRA is on that has the configured optimalAdapterSize.</p>
|
||
<p>Conceptually a slot corresponds to a r=1, 1-layer, 1-module set of in/out weights. Page width is set to the number of weights in smallest module.</p>
|
||
<p>The number of slots per page is then ceilDiv(num weights in optimally sized LoRA, num weights in smallest module)</p>
|
||
<p>Cache pages are allocated on one or more blocks </p>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a52efeeb85da595056495553753dd5170"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache10TaskIdTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache10TaskIdTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache10TaskIdTypeE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a1db70e10833cee56b07d5d2c6a8895ef"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TaskIdType</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">uint64_t</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache10TaskIdTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache28TaskLayerModuleConfigListPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache28TaskLayerModuleConfigListPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache28TaskLayerModuleConfigListPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a806bd28154016a6d2f81ccdf676b3ebb"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TaskLayerModuleConfigListPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfigE" title="tensorrt_llm::runtime::LoraCache::TaskLayerModuleConfig"><span class="n"><span class="pre">TaskLayerModuleConfig</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache28TaskLayerModuleConfigListPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache9LoraCacheERK26LoraCachePageManagerConfigRK11ModelConfigRK11WorldConfigRK13BufferManager">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache9LoraCacheERK26LoraCachePageManagerConfigRK11ModelConfigRK11WorldConfigRK13BufferManager"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache9LoraCacheERK26LoraCachePageManagerConfigRK11ModelConfigRK11WorldConfigRK13BufferManager"></span><span id="tensorrt_llm::runtime::LoraCache::LoraCache__LoraCachePageManagerConfigCR.ModelConfigCR.WorldConfigCR.BufferManagerCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1ae8c4d09102e5af60a7b48f4657db3431"></span><span class="sig-name descname"><span class="n"><span class="pre">LoraCache</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfigE" title="tensorrt_llm::runtime::LoraCachePageManagerConfig"><span class="n"><span class="pre">LoraCachePageManagerConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">pageManagerConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE" title="tensorrt_llm::runtime::ModelConfig"><span class="n"><span class="pre">ModelConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfigE" title="tensorrt_llm::runtime::WorldConfig"><span class="n"><span class="pre">WorldConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">worldConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManagerE" title="tensorrt_llm::runtime::BufferManager"><span class="n"><span class="pre">BufferManager</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">bufferManager</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9LoraCacheERK26LoraCachePageManagerConfigRK11ModelConfigRK11WorldConfigRK13BufferManager" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>param[in] pageManagerConfig: a <a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig"><span class="std std-ref">LoraCachePageManagerConfig</span></a> param[in] modelConfig: a <a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1ModelConfig"><span class="std std-ref">ModelConfig</span></a> param[in] worldConfig: a <a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1WorldConfig"><span class="std std-ref">WorldConfig</span></a> param[in] bufferManager: a <a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1BufferManager"><span class="std std-ref">BufferManager</span></a> only used to allocate page blocks </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache3putE10TaskIdType9TensorPtr9TensorPtrb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache3putE10TaskIdType9TensorPtr9TensorPtrb"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache3putE10TaskIdType9TensorPtr9TensorPtrb"></span><span id="tensorrt_llm::runtime::LoraCache::put__TaskIdType.TensorPtr.TensorPtr.b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a7a325af2ed20987b2248ec9155da6474"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">put</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache10TaskIdTypeE" title="tensorrt_llm::runtime::LoraCache::TaskIdType"><span class="n"><span class="pre">TaskIdType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">taskId</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TensorPtrE" title="tensorrt_llm::runtime::LoraCache::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">weights</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TensorPtrE" title="tensorrt_llm::runtime::LoraCache::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">config</span></span>, <span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">load</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">true</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache3putE10TaskIdType9TensorPtr9TensorPtrb" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>put a task in the cache, and claim pages for it, and optionally load task weights. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><ul class="simple">
|
||
<li><p><strong>taskId</strong> – <strong>[in]</strong> the task id </p></li>
|
||
<li><p><strong>weights</strong> – <strong>[in]</strong> lora weights tensor </p></li>
|
||
<li><p><strong>config</strong> – <strong>[in]</strong> lora config tensor </p></li>
|
||
<li><p><strong>load</strong> – <strong>[in]</strong> if true load weights before returning, otherwise do not </p></li>
|
||
</ul>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache11loadWeightsE10TaskIdType9TensorPtr9TensorPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache11loadWeightsE10TaskIdType9TensorPtr9TensorPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache11loadWeightsE10TaskIdType9TensorPtr9TensorPtr"></span><span id="tensorrt_llm::runtime::LoraCache::loadWeights__TaskIdType.TensorPtr.TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a0153f9e24cf6ddd13aa2b33276b0d539"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">loadWeights</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache10TaskIdTypeE" title="tensorrt_llm::runtime::LoraCache::TaskIdType"><span class="n"><span class="pre">TaskIdType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">taskId</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TensorPtrE" title="tensorrt_llm::runtime::LoraCache::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">weights</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TensorPtrE" title="tensorrt_llm::runtime::LoraCache::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">config</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache11loadWeightsE10TaskIdType9TensorPtr9TensorPtr" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>load task weights. This method must be called after put. It is designed to be called asynchronously after put returns with load = false </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><ul class="simple">
|
||
<li><p><strong>taslId</strong> – <strong>[in]</strong> the task id </p></li>
|
||
<li><p><strong>weights</strong> – <strong>[in]</strong> lora weights tensor </p></li>
|
||
<li><p><strong>config</strong> – <strong>[in]</strong> lora config tensor </p></li>
|
||
</ul>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime9LoraCache8isLoadedE10TaskIdType">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime9LoraCache8isLoadedE10TaskIdType"></span><span id="_CPPv2NK12tensorrt_llm7runtime9LoraCache8isLoadedE10TaskIdType"></span><span id="tensorrt_llm::runtime::LoraCache::isLoaded__TaskIdTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a0c328425bd4342282f2e948f62fa052b"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">isLoaded</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache10TaskIdTypeE" title="tensorrt_llm::runtime::LoraCache::TaskIdType"><span class="n"><span class="pre">TaskIdType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">taskId</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime9LoraCache8isLoadedE10TaskIdType" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>taskId</strong> – <strong>[in]</strong> the task id </p>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>&#8212; true if task is loaded (weights are in place) and false otherwise </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime9LoraCache6isDoneE10TaskIdType">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime9LoraCache6isDoneE10TaskIdType"></span><span id="_CPPv2NK12tensorrt_llm7runtime9LoraCache6isDoneE10TaskIdType"></span><span id="tensorrt_llm::runtime::LoraCache::isDone__TaskIdTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a4ef43f6ccf206bb0c82a53f2d6474d0e"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">isDone</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache10TaskIdTypeE" title="tensorrt_llm::runtime::LoraCache::TaskIdType"><span class="n"><span class="pre">TaskIdType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">taskId</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime9LoraCache6isDoneE10TaskIdType" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>taskId</strong> – <strong>[in]</strong> the task id </p>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>&#8212; true if task is marked done and can be evicted </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime9LoraCache3hasE10TaskIdType">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime9LoraCache3hasE10TaskIdType"></span><span id="_CPPv2NK12tensorrt_llm7runtime9LoraCache3hasE10TaskIdType"></span><span id="tensorrt_llm::runtime::LoraCache::has__TaskIdTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a8baa9098fbc9800417b3efea90f27f47"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">has</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache10TaskIdTypeE" title="tensorrt_llm::runtime::LoraCache::TaskIdType"><span class="n"><span class="pre">TaskIdType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">taskId</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime9LoraCache3hasE10TaskIdType" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>taskId</strong> – <strong>[in]</strong> the task id </p>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>&#8212; true if task is in the cache (not necessarily loaded) and false otherwise </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache3getE10TaskIdType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache3getE10TaskIdType"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache3getE10TaskIdType"></span><span id="tensorrt_llm::runtime::LoraCache::get__TaskIdType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a54af79acf17beb6ecc80dc9407ee087d"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfigE" title="tensorrt_llm::runtime::LoraCache::TaskLayerModuleConfig"><span class="n"><span class="pre">TaskLayerModuleConfig</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="n"><span class="pre">get</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache10TaskIdTypeE" title="tensorrt_llm::runtime::LoraCache::TaskIdType"><span class="n"><span class="pre">TaskIdType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">taskId</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache3getE10TaskIdType" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>taskId</strong> – <strong>[in]</strong> the task id </p>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>&#8212; list of Value objects with pointers to task weights </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache4bumpE10TaskIdType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache4bumpE10TaskIdType"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache4bumpE10TaskIdType"></span><span id="tensorrt_llm::runtime::LoraCache::bump__TaskIdType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a48bc915d611d849249f459333c8a0ca6"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">bump</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache10TaskIdTypeE" title="tensorrt_llm::runtime::LoraCache::TaskIdType"><span class="n"><span class="pre">TaskIdType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">taskId</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache4bumpE10TaskIdType" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>bump task and make it the most recently used </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>taskId</strong> – <strong>[in]</strong> the task id </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache12markTaskDoneE10TaskIdType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache12markTaskDoneE10TaskIdType"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache12markTaskDoneE10TaskIdType"></span><span id="tensorrt_llm::runtime::LoraCache::markTaskDone__TaskIdType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1ac753d065a24dcee086f6d64825dc1249"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">markTaskDone</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache10TaskIdTypeE" title="tensorrt_llm::runtime::LoraCache::TaskIdType"><span class="n"><span class="pre">TaskIdType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">taskId</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache12markTaskDoneE10TaskIdType" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>mark task done meaning it can be evicted </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>taskId</strong> – <strong>[in]</strong> the task id </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache11markAllDoneEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache11markAllDoneEv"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache11markAllDoneEv"></span><span id="tensorrt_llm::runtime::LoraCache::markAllDone"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1ad6fb2ca1f0ab34248ebed9df893907d0"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">markAllDone</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache11markAllDoneEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>mark all tasks in cache done </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime9LoraCache17determineNumPagesE10TaskIdType">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime9LoraCache17determineNumPagesE10TaskIdType"></span><span id="_CPPv2NK12tensorrt_llm7runtime9LoraCache17determineNumPagesE10TaskIdType"></span><span id="tensorrt_llm::runtime::LoraCache::determineNumPages__TaskIdTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a4690a3d893c213396d34fe1c4242dcf6"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">determineNumPages</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache10TaskIdTypeE" title="tensorrt_llm::runtime::LoraCache::TaskIdType"><span class="n"><span class="pre">TaskIdType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">taskId</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime9LoraCache17determineNumPagesE10TaskIdType" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>taskId</strong> – <strong>[in]</strong> the taskid </p>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>&#8212; number of pages needed to store the given task </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime9LoraCache17determineNumPagesE9TensorPtr">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime9LoraCache17determineNumPagesE9TensorPtr"></span><span id="_CPPv2NK12tensorrt_llm7runtime9LoraCache17determineNumPagesE9TensorPtr"></span><span id="tensorrt_llm::runtime::LoraCache::determineNumPages__TensorPtrC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1af8e36fb16dcc5ecb114b6bd1e4f89b60"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">determineNumPages</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TensorPtrE" title="tensorrt_llm::runtime::LoraCache::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">config</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime9LoraCache17determineNumPagesE9TensorPtr" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>config</strong> – <strong>[in]</strong> lora config tensor </p>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>&#8212; number of pages needed to store the task configured with config tensor </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime9LoraCache4fitsE9TensorPtr">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime9LoraCache4fitsE9TensorPtr"></span><span id="_CPPv2NK12tensorrt_llm7runtime9LoraCache4fitsE9TensorPtr"></span><span id="tensorrt_llm::runtime::LoraCache::fits__TensorPtrC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1aee17e54b1070ffca531aaf6ff9fdfdef"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">fits</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TensorPtrE" title="tensorrt_llm::runtime::LoraCache::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">config</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime9LoraCache4fitsE9TensorPtr" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>config</strong> – <strong>[in]</strong> a lora config tensor </p>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>&#8212; true in task fits in cache false otherwise </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache8copyTaskE10TaskIdTypeR9LoraCacheb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache8copyTaskE10TaskIdTypeR9LoraCacheb"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache8copyTaskE10TaskIdTypeR9LoraCacheb"></span><span id="tensorrt_llm::runtime::LoraCache::copyTask__TaskIdType.LoraCacheR.b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a3fe38aa85a89b76a354cca719b65d39b"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">copyTask</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache10TaskIdTypeE" title="tensorrt_llm::runtime::LoraCache::TaskIdType"><span class="n"><span class="pre">TaskIdType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">taskId</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCacheE" title="tensorrt_llm::runtime::LoraCache"><span class="n"><span class="pre">LoraCache</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">deviceCache</span></span>, <span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">markDone</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">false</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache8copyTaskE10TaskIdTypeR9LoraCacheb" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>copy task to another cache. Caches must have the same page size. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><ul class="simple">
|
||
<li><p><strong>taskId</strong> – <strong>[in]</strong> the task id to copy </p></li>
|
||
<li><p><strong>otherCache</strong> – <strong>[in]</strong> the <a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1LoraCache"><span class="std std-ref">LoraCache</span></a> to move the task to </p></li>
|
||
<li><p><strong>markDone</strong> – <strong>[in]</strong> mark the copied task done as it’s copied </p></li>
|
||
</ul>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime9LoraCache11getNumPagesEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime9LoraCache11getNumPagesEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime9LoraCache11getNumPagesEv"></span><span id="tensorrt_llm::runtime::LoraCache::getNumPagesC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1adcd24dca8d65f4862021ae2fa84b9199"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getNumPages</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime9LoraCache11getNumPagesEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>&#8212; total number of pages allocated to cache (used or not) </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime9LoraCache10getPagePtrE6size_t">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime9LoraCache10getPagePtrE6size_t"></span><span id="_CPPv2NK12tensorrt_llm7runtime9LoraCache10getPagePtrE6size_t"></span><span id="tensorrt_llm::runtime::LoraCache::getPagePtr__sC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1af4191f7de099ddabb12699b7c7613812"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor14SharedConstPtrE" title="tensorrt_llm::runtime::ITensor::SharedConstPtr"><span class="n"><span class="pre">SharedConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getPagePtr</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">pageId</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime9LoraCache10getPagePtrE6size_t" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>pageId</strong> – <strong>[in]</strong> the page id </p>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>&#8212; const pointer to page </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-functions">Public Static Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache11copyToPagesE9TensorPtr9TensorPtrRK11ModelConfigRK11WorldConfigNSt13unordered_mapI10SizeType3210LoraModuleEERK13BufferManagerRKNSt6vectorI9TensorPtrEERKNSt6vectorINSt6size_tEEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache11copyToPagesE9TensorPtr9TensorPtrRK11ModelConfigRK11WorldConfigNSt13unordered_mapI10SizeType3210LoraModuleEERK13BufferManagerRKNSt6vectorI9TensorPtrEERKNSt6vectorINSt6size_tEEE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache11copyToPagesE9TensorPtr9TensorPtrRK11ModelConfigRK11WorldConfigNSt13unordered_mapI10SizeType3210LoraModuleEERK13BufferManagerRKNSt6vectorI9TensorPtrEERKNSt6vectorINSt6size_tEEE"></span><span id="tensorrt_llm::runtime::LoraCache::copyToPages__TensorPtr.TensorPtr.ModelConfigCR.WorldConfigCR.std::unordered_map:SizeType32.LoraModule:.BufferManagerCR.std::vector:TensorPtr:CR.std::vector:std::s:CR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1ac67eb5e9fdcb057e560bdb2e62e89ef5"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCacheE" title="tensorrt_llm::runtime::LoraCache"><span class="n"><span class="pre">LoraCache</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfigE" title="tensorrt_llm::runtime::LoraCache::TaskLayerModuleConfig"><span class="n"><span class="pre">TaskLayerModuleConfig</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">copyToPages</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TensorPtrE" title="tensorrt_llm::runtime::LoraCache::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">weights</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TensorPtrE" title="tensorrt_llm::runtime::LoraCache::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">config</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE" title="tensorrt_llm::runtime::ModelConfig"><span class="n"><span class="pre">ModelConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfigE" title="tensorrt_llm::runtime::WorldConfig"><span class="n"><span class="pre">WorldConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">worldConfig</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">unordered_map</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">,</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModuleE" title="tensorrt_llm::runtime::LoraModule"><span class="n"><span class="pre">LoraModule</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="n sig-param"><span class="pre">moduleIdToModel</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManagerE" title="tensorrt_llm::runtime::BufferManager"><span class="n"><span class="pre">BufferManager</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">manager</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TensorPtrE" title="tensorrt_llm::runtime::LoraCache::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">pages</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">pageIds</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache11copyToPagesE9TensorPtr9TensorPtrRK11ModelConfigRK11WorldConfigNSt13unordered_mapI10SizeType3210LoraModuleEERK13BufferManagerRKNSt6vectorI9TensorPtrEERKNSt6vectorINSt6size_tEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Copy task weights to cache pages. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><ul class="simple">
|
||
<li><p><strong>weights</strong> – <strong>[in]</strong> task weights </p></li>
|
||
<li><p><strong>config</strong> – <strong>[in]</strong> task config tensor </p></li>
|
||
<li><p><strong>modelConfig</strong> – <strong>[in]</strong> a <a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1ModelConfig"><span class="std std-ref">ModelConfig</span></a></p></li>
|
||
<li><p><strong>worldConfig</strong> – <strong>[in]</strong> a <a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1WorldConfig"><span class="std std-ref">WorldConfig</span></a></p></li>
|
||
<li><p><strong>modelIdToModel</strong> – <strong>[in]</strong> map from lora module id to <a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1LoraModule"><span class="std std-ref">LoraModule</span></a></p></li>
|
||
<li><p><strong>manager</strong> – <strong>[in]</strong> a <a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1BufferManager"><span class="std std-ref">BufferManager</span></a> the manager to use to perform the copies </p></li>
|
||
<li><p><strong>pages</strong> – <strong>[out]</strong> list of page tensors to copy weights to </p></li>
|
||
<li><p><strong>pageIds</strong> – <strong>[in]</strong> page ids for the pages </p></li>
|
||
</ul>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>&#8212; list of cache Values objects </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache17splitTransposeCpuER7ITensorRK7ITensor10SizeType3210SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache17splitTransposeCpuER7ITensorRK7ITensor10SizeType3210SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache17splitTransposeCpuER7ITensorRK7ITensor10SizeType3210SizeType32"></span><span id="tensorrt_llm::runtime::LoraCache::splitTransposeCpu__ITensorR.ITensorCR.SizeType32.SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a2fc01315185a952f749316961be48494"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">splitTransposeCpu</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">output</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">input</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tpSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tpRank</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache17splitTransposeCpuER7ITensorRK7ITensor10SizeType3210SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>splits second dim of input into tpSize parts and writes the tpRank split to output </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><ul class="simple">
|
||
<li><p><strong>output</strong> – <strong>[out]</strong> output tensor </p></li>
|
||
<li><p><strong>input</strong> – <strong>[in]</strong> input tensor </p></li>
|
||
<li><p><strong>tpSize</strong> – <strong>[in]</strong> number of splits </p></li>
|
||
<li><p><strong>tpRank</strong> – <strong>[in]</strong> the split to write to output </p></li>
|
||
</ul>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-types">Private Types</p>
|
||
<dl class="cpp enum">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache11ValueStatusE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache11ValueStatusE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache11ValueStatusE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a75949c4322be311f8f485c1b3983a4a8"></span><span class="k"><span class="pre">enum</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">ValueStatus</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache11ValueStatusE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p><em>Values:</em></p>
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache11ValueStatus21kVALUE_STATUS_MISSINGE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache11ValueStatus21kVALUE_STATUS_MISSINGE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache11ValueStatus21kVALUE_STATUS_MISSINGE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a75949c4322be311f8f485c1b3983a4a8a71919c7fd85566514a9abcd3e631e892"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kVALUE_STATUS_MISSING</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache11ValueStatus21kVALUE_STATUS_MISSINGE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache11ValueStatus24kVALUE_STATUS_PROCESSINGE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache11ValueStatus24kVALUE_STATUS_PROCESSINGE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache11ValueStatus24kVALUE_STATUS_PROCESSINGE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a75949c4322be311f8f485c1b3983a4a8a7693ca37a375c6f160c7608ecf999617"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kVALUE_STATUS_PROCESSING</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache11ValueStatus24kVALUE_STATUS_PROCESSINGE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache11ValueStatus20kVALUE_STATUS_LOADEDE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache11ValueStatus20kVALUE_STATUS_LOADEDE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache11ValueStatus20kVALUE_STATUS_LOADEDE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a75949c4322be311f8f485c1b3983a4a8a599f5d302d900314903fd3ab8c2be0d1"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kVALUE_STATUS_LOADED</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache11ValueStatus20kVALUE_STATUS_LOADEDE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache12TaskValuePtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache12TaskValuePtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache12TaskValuePtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1af49013c48c41ec7b89e862a35f36bbee"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TaskValuePtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValueE" title="tensorrt_llm::runtime::LoraCache::TaskValue"><span class="n"><span class="pre">TaskValue</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache12TaskValuePtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-functions">Private Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache11loadWeightsER9TaskValue9TensorPtr9TensorPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache11loadWeightsER9TaskValue9TensorPtr9TensorPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache11loadWeightsER9TaskValue9TensorPtr9TensorPtr"></span><span id="tensorrt_llm::runtime::LoraCache::loadWeights__TaskValueR.TensorPtr.TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a0264725d06825eee042d00c976c20cf6"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">loadWeights</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValueE" title="tensorrt_llm::runtime::LoraCache::TaskValue"><span class="n"><span class="pre">TaskValue</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">cacheValue</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TensorPtrE" title="tensorrt_llm::runtime::LoraCache::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">weights</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TensorPtrE" title="tensorrt_llm::runtime::LoraCache::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">config</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache11loadWeightsER9TaskValue9TensorPtr9TensorPtr" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache18bumpTaskInProgressE10TaskIdType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache18bumpTaskInProgressE10TaskIdType"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache18bumpTaskInProgressE10TaskIdType"></span><span id="tensorrt_llm::runtime::LoraCache::bumpTaskInProgress__TaskIdType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a776bdc112fb5c1d966d83d8c5fbe0f40"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">bumpTaskInProgress</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache10TaskIdTypeE" title="tensorrt_llm::runtime::LoraCache::TaskIdType"><span class="n"><span class="pre">TaskIdType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">taskId</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache18bumpTaskInProgressE10TaskIdType" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime9LoraCache9getStatusE10TaskIdType">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime9LoraCache9getStatusE10TaskIdType"></span><span id="_CPPv2NK12tensorrt_llm7runtime9LoraCache9getStatusE10TaskIdType"></span><span id="tensorrt_llm::runtime::LoraCache::getStatus__TaskIdTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1aa0f3c256a9ebc2d30465a6a9e490fb42"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache11ValueStatusE" title="tensorrt_llm::runtime::LoraCache::ValueStatus"><span class="n"><span class="pre">ValueStatus</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getStatus</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache10TaskIdTypeE" title="tensorrt_llm::runtime::LoraCache::TaskIdType"><span class="n"><span class="pre">TaskIdType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">taskId</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime9LoraCache9getStatusE10TaskIdType" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache19claimPagesWithEvictE10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache19claimPagesWithEvictE10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache19claimPagesWithEvictE10SizeType32"></span><span id="tensorrt_llm::runtime::LoraCache::claimPagesWithEvict__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a42cd991f5baeba5c72aff9d90fe2932d"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">claimPagesWithEvict</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">numPages</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache19claimPagesWithEvictE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>claim numPages, evicting tasks if needed </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>numPages</strong> – <strong>[in]</strong> number of pages to claim </p>
|
||
</dd>
|
||
<dt class="field-even">Throws<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p><span><span class="cpp-expr sig sig-inline cpp"><span class="n">std</span><span class="p">::</span><span class="n">runtime_error</span></span></span> – if all pages cannot be claimed </p>
|
||
</dd>
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>&#8212; list of page ids </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache16copyTaskMapPagesER9TaskValueRK9TaskValueRKNSt6vectorI6size_tEERK9LoraCache">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache16copyTaskMapPagesER9TaskValueRK9TaskValueRKNSt6vectorI6size_tEERK9LoraCache"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache16copyTaskMapPagesER9TaskValueRK9TaskValueRKNSt6vectorI6size_tEERK9LoraCache"></span><span id="tensorrt_llm::runtime::LoraCache::copyTaskMapPages__TaskValueR.TaskValueCR.std::vector:s:CR.LoraCacheCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a3cd51bc551695aade65b9683f2ed622c"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">map</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">size_t</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">pair</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">size_t</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">copyTaskMapPages</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValueE" title="tensorrt_llm::runtime::LoraCache::TaskValue"><span class="n"><span class="pre">TaskValue</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">targetTaskValue</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValueE" title="tensorrt_llm::runtime::LoraCache::TaskValue"><span class="n"><span class="pre">TaskValue</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">sourceTaskValue</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">size_t</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">targetPageIds</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCacheE" title="tensorrt_llm::runtime::LoraCache"><span class="n"><span class="pre">LoraCache</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">targetCache</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache16copyTaskMapPagesER9TaskValueRK9TaskValueRKNSt6vectorI6size_tEERK9LoraCache" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Internal helper method used inside copyTask. Not thread safe on its own </p>
|
||
</dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-members">Private Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache18mPageManagerConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache18mPageManagerConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache18mPageManagerConfigE"></span><span id="tensorrt_llm::runtime::LoraCache::mPageManagerConfig__LoraCachePageManagerConfig"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a48ca80214b9115983030b1b406c1e4fc"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfigE" title="tensorrt_llm::runtime::LoraCachePageManagerConfig"><span class="n"><span class="pre">LoraCachePageManagerConfig</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mPageManagerConfig</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache18mPageManagerConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache12mModelConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache12mModelConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache12mModelConfigE"></span><span id="tensorrt_llm::runtime::LoraCache::mModelConfig__ModelConfig"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a0064ca2bd7e388ef4766534041f0d867"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE" title="tensorrt_llm::runtime::ModelConfig"><span class="n"><span class="pre">ModelConfig</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mModelConfig</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache12mModelConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache12mWorldConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache12mWorldConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache12mWorldConfigE"></span><span id="tensorrt_llm::runtime::LoraCache::mWorldConfig__WorldConfig"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a85f9991a8aa4f25db6d6b3505474f6c9"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfigE" title="tensorrt_llm::runtime::WorldConfig"><span class="n"><span class="pre">WorldConfig</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mWorldConfig</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache12mWorldConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache11mPagesMutexE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache11mPagesMutexE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache11mPagesMutexE"></span><span id="tensorrt_llm::runtime::LoraCache::mPagesMutex__std::mutex"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1acced4dd499010314b3ba5c6661e1f4cd"></span><span class="k"><span class="pre">mutable</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">mutex</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mPagesMutex</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache11mPagesMutexE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache17mCachePageManagerE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache17mCachePageManagerE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache17mCachePageManagerE"></span><span id="tensorrt_llm::runtime::LoraCache::mCachePageManager__std::unique_ptr:LoraCachePageManager:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a7a349c6ef0735c024bd59e60f3a2ff62"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">unique_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime20LoraCachePageManagerE" title="tensorrt_llm::runtime::LoraCachePageManager"><span class="n"><span class="pre">LoraCachePageManager</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mCachePageManager</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache17mCachePageManagerE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache11mCacheMutexE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache11mCacheMutexE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache11mCacheMutexE"></span><span id="tensorrt_llm::runtime::LoraCache::mCacheMutex__std::mutex"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1ad05d0ba4aa25bffef6f40d88119e4a4b"></span><span class="k"><span class="pre">mutable</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">mutex</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mCacheMutex</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache11mCacheMutexE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache9mCacheMapE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache9mCacheMapE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache9mCacheMapE"></span><span id="tensorrt_llm::runtime::LoraCache::mCacheMap__std::unordered_map:TaskIdType.TaskValuePtr:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a925da94c9416d9e1206c5a79bd772c60"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">unordered_map</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache10TaskIdTypeE" title="tensorrt_llm::runtime::LoraCache::TaskIdType"><span class="n"><span class="pre">TaskIdType</span></span></a><span class="p"><span class="pre">,</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache12TaskValuePtrE" title="tensorrt_llm::runtime::LoraCache::TaskValuePtr"><span class="n"><span class="pre">TaskValuePtr</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mCacheMap</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9mCacheMapE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache16mInProgressTasksE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache16mInProgressTasksE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache16mInProgressTasksE"></span><span id="tensorrt_llm::runtime::LoraCache::mInProgressTasks__std::list:TaskIdType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a0f51387f4a795d072b2c53389c2ad0e7"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">list</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache10TaskIdTypeE" title="tensorrt_llm::runtime::LoraCache::TaskIdType"><span class="n"><span class="pre">TaskIdType</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mInProgressTasks</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache16mInProgressTasksE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache10mDoneTasksE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache10mDoneTasksE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache10mDoneTasksE"></span><span id="tensorrt_llm::runtime::LoraCache::mDoneTasks__std::list:TaskIdType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a7786c6df0ce94dc5cbc7f877d989c932"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">list</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache10TaskIdTypeE" title="tensorrt_llm::runtime::LoraCache::TaskIdType"><span class="n"><span class="pre">TaskIdType</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mDoneTasks</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache10mDoneTasksE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache21mDeviceBufferManagersE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache21mDeviceBufferManagersE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache21mDeviceBufferManagersE"></span><span id="tensorrt_llm::runtime::LoraCache::mDeviceBufferManagers__std::vector:std::unique_ptr:BufferManager::"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1aaee288e322ff0875437dc466b889b605"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">unique_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManagerE" title="tensorrt_llm::runtime::BufferManager"><span class="n"><span class="pre">BufferManager</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mDeviceBufferManagers</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache21mDeviceBufferManagersE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache14mBufferManagerE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache14mBufferManagerE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache14mBufferManagerE"></span><span id="tensorrt_llm::runtime::LoraCache::mBufferManager__std::unique_ptr:BufferManager:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a3b33e7cb854d144ec9f62c0724dcfeb1"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">unique_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManagerE" title="tensorrt_llm::runtime::BufferManager"><span class="n"><span class="pre">BufferManager</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mBufferManager</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache14mBufferManagerE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache17mModuleIdToModuleE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache17mModuleIdToModuleE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache17mModuleIdToModuleE"></span><span id="tensorrt_llm::runtime::LoraCache::mModuleIdToModule__std::unordered_map:SizeType32.LoraModule:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a57ff4ff69308b55880fc4460dd1d4a4f"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">unordered_map</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">,</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModuleE" title="tensorrt_llm::runtime::LoraModule"><span class="n"><span class="pre">LoraModule</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mModuleIdToModule</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache17mModuleIdToModuleE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-static-functions">Private Static Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime9LoraCache22splitTransposeCpuInnerEvR7ITensorRK7ITensor10SizeType3210SizeType32">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime9LoraCache22splitTransposeCpuInnerER7ITensorRK7ITensor10SizeType3210SizeType32"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime9LoraCache22splitTransposeCpuInnerER7ITensorRK7ITensor10SizeType3210SizeType32"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCache_1a5ebe7179a3661f5e3ada638cd0fc578a"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">splitTransposeCpuInner</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">output</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">input</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tpSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tpRank</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime9LoraCache22splitTransposeCpuInnerEvR7ITensorRK7ITensor10SizeType3210SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfigE"></span><span id="tensorrt_llm::runtime::LoraCache::TaskLayerModuleConfig"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1LoraCache_1_1TaskLayerModuleConfig"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TaskLayerModuleConfig</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="docutils container">
|
||
<em>#include <loraCache.h></em></div>
|
||
<p>Contains information on a single layer / module. A list of these configs is associated with each task and can be used to populate runtime tensors. </p>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig8toStringEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig8toStringEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig8toStringEv"></span><span id="tensorrt_llm::runtime::LoraCache::TaskLayerModuleConfig::toStringC"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1LoraCache_1_1TaskLayerModuleConfig_1a22d58d87f683ff4c3c110bcea6ac76cf"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">toString</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig8toStringEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfigeqERKN9LoraCache21TaskLayerModuleConfigE">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfigeqERKN9LoraCache21TaskLayerModuleConfigE"></span><span id="_CPPv2NK12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfigeqERKN9LoraCache21TaskLayerModuleConfigE"></span><span id="tensorrt_llm::runtime::LoraCache::TaskLayerModuleConfig::eq-operator__LoraCache::TaskLayerModuleConfigCRC"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1LoraCache_1_1TaskLayerModuleConfig_1aeeb90d000b1ebe5532ec51e9012b3d73"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="k"><span class="pre">operator</span></span><span class="o"><span class="pre">==</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCacheE" title="tensorrt_llm::runtime::LoraCache"><span class="n"><span class="pre">LoraCache</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfigE" title="tensorrt_llm::runtime::LoraCache::TaskLayerModuleConfig"><span class="n"><span class="pre">TaskLayerModuleConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">o</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfigeqERKN9LoraCache21TaskLayerModuleConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig6pageIdE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig6pageIdE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig6pageIdE"></span><span id="tensorrt_llm::runtime::LoraCache::TaskLayerModuleConfig::pageId__std::s"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1LoraCache_1_1TaskLayerModuleConfig_1acb73dea546a2d4be0143cb0f01a81c15"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">pageId</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig6pageIdE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig7slotIdxE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig7slotIdxE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig7slotIdxE"></span><span id="tensorrt_llm::runtime::LoraCache::TaskLayerModuleConfig::slotIdx__SizeType32"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1LoraCache_1_1TaskLayerModuleConfig_1a7ab99965797b2b31e63b66bab9c7dd9c"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">slotIdx</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig7slotIdxE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig6inSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig6inSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig6inSizeE"></span><span id="tensorrt_llm::runtime::LoraCache::TaskLayerModuleConfig::inSize__SizeType32"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1LoraCache_1_1TaskLayerModuleConfig_1a2cbe1e0f50a01f3d781216c627c6b0ba"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">inSize</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig6inSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig7outSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig7outSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig7outSizeE"></span><span id="tensorrt_llm::runtime::LoraCache::TaskLayerModuleConfig::outSize__SizeType32"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1LoraCache_1_1TaskLayerModuleConfig_1a2ff87ee039d5c46915d981c11114ddf0"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">outSize</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig7outSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig8moduleIdE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig8moduleIdE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig8moduleIdE"></span><span id="tensorrt_llm::runtime::LoraCache::TaskLayerModuleConfig::moduleId__SizeType32"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1LoraCache_1_1TaskLayerModuleConfig_1ab0a616711dea88ac63fe31043699b0b9"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">moduleId</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig8moduleIdE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig7layerIdE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig7layerIdE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig7layerIdE"></span><span id="tensorrt_llm::runtime::LoraCache::TaskLayerModuleConfig::layerId__SizeType32"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1LoraCache_1_1TaskLayerModuleConfig_1a02012816d82515dd66bad77eb9e0ea26"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">layerId</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig7layerIdE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig11adapterSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig11adapterSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig11adapterSizeE"></span><span id="tensorrt_llm::runtime::LoraCache::TaskLayerModuleConfig::adapterSize__SizeType32"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1LoraCache_1_1TaskLayerModuleConfig_1af91dca0e1a714f02cd9d62979ef78463"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">adapterSize</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig11adapterSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig8numSlotsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig8numSlotsE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig8numSlotsE"></span><span id="tensorrt_llm::runtime::LoraCache::TaskLayerModuleConfig::numSlots__SizeType32"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1LoraCache_1_1TaskLayerModuleConfig_1a0b6d002a8a14a64f26b54152b614ab3b"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">numSlots</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig8numSlotsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig16weightsInPointerE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig16weightsInPointerE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig16weightsInPointerE"></span><span id="tensorrt_llm::runtime::LoraCache::TaskLayerModuleConfig::weightsInPointer__std::int64_t"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1LoraCache_1_1TaskLayerModuleConfig_1a67ab9af3fd4782431150b26241f49cdc"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">int64_t</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">weightsInPointer</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig16weightsInPointerE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig17weightsOutPointerE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig17weightsOutPointerE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig17weightsOutPointerE"></span><span id="tensorrt_llm::runtime::LoraCache::TaskLayerModuleConfig::weightsOutPointer__std::int64_t"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1LoraCache_1_1TaskLayerModuleConfig_1aa5883002221d6f5771219423d5aadb14"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">int64_t</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">weightsOutPointer</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache21TaskLayerModuleConfig17weightsOutPointerE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-friends">Friends</p>
|
||
<dl>
|
||
<dt class="sig sig-object cpp">
|
||
<em class="property"><span class="pre">friend</span> <span class="pre">class</span></em> <span class="pre">TaskLayerModuleConfigBindings</span></dt>
|
||
</dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValueE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache9TaskValueE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache9TaskValueE"></span><span id="tensorrt_llm::runtime::LoraCache::TaskValue"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1LoraCache_1_1TaskValue"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TaskValue</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValueE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Holds configuration and state for a single task. </p>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValue9TaskValueEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache9TaskValue9TaskValueEv"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache9TaskValue9TaskValueEv"></span><span id="tensorrt_llm::runtime::LoraCache::TaskValue::TaskValue"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1LoraCache_1_1TaskValue_1ac50c3976697f9adbf75af8f5ff4398ad"></span><span class="sig-name descname"><span class="n"><span class="pre">TaskValue</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">delete</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValue9TaskValueEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValueD0Ev">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache9TaskValueD0Ev"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache9TaskValueD0Ev"></span><span id="tensorrt_llm::runtime::LoraCache::TaskValue::~TaskValue"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1LoraCache_1_1TaskValue_1a03e4986e9975fe5f53eae770fc730611"></span><span class="sig-name descname"><span class="n"><span class="pre">~TaskValue</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">default</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValueD0Ev" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValue9TaskValueERKNSt6vectorINSt6size_tEEERK28TaskLayerModuleConfigListPtrNSt4listI10TaskIdTypeE8iteratorEbbbb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache9TaskValue9TaskValueERKNSt6vectorINSt6size_tEEERK28TaskLayerModuleConfigListPtrNSt4listI10TaskIdTypeE8iteratorEbbbb"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache9TaskValue9TaskValueERKNSt6vectorINSt6size_tEEERK28TaskLayerModuleConfigListPtrNSt4listI10TaskIdTypeE8iteratorEbbbb"></span><span id="tensorrt_llm::runtime::LoraCache::TaskValue::TaskValue__std::vector:std::s:CR.TaskLayerModuleConfigListPtrCR.std::list:TaskIdType:::iterator.b.b.b.b"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1LoraCache_1_1TaskValue_1a57e05d4fc10716c2ace0d4cd5373be2a"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TaskValue</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">pageIds</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache28TaskLayerModuleConfigListPtrE" title="tensorrt_llm::runtime::LoraCache::TaskLayerModuleConfigListPtr"><span class="n"><span class="pre">TaskLayerModuleConfigListPtr</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">configs</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">list</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache10TaskIdTypeE" title="tensorrt_llm::runtime::LoraCache::TaskIdType"><span class="n"><span class="pre">TaskIdType</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">iterator</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">it</span></span>, <span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">inProgress</span></span>, <span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">loaded</span></span>, <span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">done</span></span>, <span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">loadInProgress</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">false</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValue9TaskValueERKNSt6vectorINSt6size_tEEERK28TaskLayerModuleConfigListPtrNSt4listI10TaskIdTypeE8iteratorEbbbb" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValue9TaskValueERR9TaskValue">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache9TaskValue9TaskValueERR9TaskValue"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache9TaskValue9TaskValueERR9TaskValue"></span><span id="tensorrt_llm::runtime::LoraCache::TaskValue::TaskValue__TaskValueRR"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1LoraCache_1_1TaskValue_1a5691f27c3933372b45897851e372830e"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TaskValue</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValue9TaskValueERR9TaskValue" title="tensorrt_llm::runtime::LoraCache::TaskValue::TaskValue"><span class="n"><span class="pre">TaskValue</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">o</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValue9TaskValueERR9TaskValue" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValueaSERR9TaskValue">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache9TaskValueaSERR9TaskValue"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache9TaskValueaSERR9TaskValue"></span><span id="tensorrt_llm::runtime::LoraCache::TaskValue::assign-operator__TaskValueRR"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1LoraCache_1_1TaskValue_1a347d57780454a36ce09ea22e211a08d9"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValueE" title="tensorrt_llm::runtime::LoraCache::TaskValue"><span class="n"><span class="pre">TaskValue</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="k"><span class="pre">operator</span></span><span class="o"><span class="pre">=</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValueE" title="tensorrt_llm::runtime::LoraCache::TaskValue"><span class="n"><span class="pre">TaskValue</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">o</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValueaSERR9TaskValue" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValue7pageIdsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache9TaskValue7pageIdsE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache9TaskValue7pageIdsE"></span><span id="tensorrt_llm::runtime::LoraCache::TaskValue::pageIds__std::vector:std::s:"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1LoraCache_1_1TaskValue_1a374f2e22fe8185f13d0900cba0930828"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">pageIds</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValue7pageIdsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValue7configsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache9TaskValue7configsE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache9TaskValue7configsE"></span><span id="tensorrt_llm::runtime::LoraCache::TaskValue::configs__TaskLayerModuleConfigListPtr"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1LoraCache_1_1TaskValue_1a55350722530644cc428078c07202defc"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache28TaskLayerModuleConfigListPtrE" title="tensorrt_llm::runtime::LoraCache::TaskLayerModuleConfigListPtr"><span class="n"><span class="pre">TaskLayerModuleConfigListPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">configs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValue7configsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValue2itE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache9TaskValue2itE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache9TaskValue2itE"></span><span id="tensorrt_llm::runtime::LoraCache::TaskValue::it__std::list:TaskIdType:::iterator"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1LoraCache_1_1TaskValue_1a86d754491bf621bea52a641aa7f2c05a"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">list</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache10TaskIdTypeE" title="tensorrt_llm::runtime::LoraCache::TaskIdType"><span class="n"><span class="pre">TaskIdType</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">iterator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">it</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValue2itE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValue10inProgressE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache9TaskValue10inProgressE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache9TaskValue10inProgressE"></span><span id="tensorrt_llm::runtime::LoraCache::TaskValue::inProgress__b"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1LoraCache_1_1TaskValue_1a75db6091b5e8ffe849dcb99f0b50c383"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">inProgress</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValue10inProgressE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValue6loadedE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache9TaskValue6loadedE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache9TaskValue6loadedE"></span><span id="tensorrt_llm::runtime::LoraCache::TaskValue::loaded__b"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1LoraCache_1_1TaskValue_1a3f073e2982f8d32b58fd2836ac1cea96"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">loaded</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValue6loadedE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValue4doneE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache9TaskValue4doneE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache9TaskValue4doneE"></span><span id="tensorrt_llm::runtime::LoraCache::TaskValue::done__b"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1LoraCache_1_1TaskValue_1aab94235b5a3c2d5ab7187bbbe334d57e"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">done</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValue4doneE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Marks a task a done. This is used to mark a task as done during loading. if done=true at the end of loading (end of put, loadweights, or copyTask) the task will be marked as done </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValue14loadInProgressE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9LoraCache9TaskValue14loadInProgressE"></span><span id="_CPPv2N12tensorrt_llm7runtime9LoraCache9TaskValue14loadInProgressE"></span><span id="tensorrt_llm::runtime::LoraCache::TaskValue::loadInProgress__b"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1LoraCache_1_1TaskValue_1a98f1c9b6a5fe00e9009b366bf3861a17"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">loadInProgress</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9LoraCache9TaskValue14loadInProgressE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Indicates weights are loading either in put or loadWeights This is used to block concurrent loadWeights calls for the same task. </p>
|
||
</dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime22LoraCacheFullExceptionE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime22LoraCacheFullExceptionE"></span><span id="_CPPv2N12tensorrt_llm7runtime22LoraCacheFullExceptionE"></span><span id="tensorrt_llm::runtime::LoraCacheFullException"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCacheFullException"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">LoraCacheFullException</span></span></span><span class="w"> </span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="k"><span class="pre">public</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime21LoraExpectedExceptionE" title="tensorrt_llm::runtime::LoraExpectedException"><span class="n"><span class="pre">LoraExpectedException</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime22LoraCacheFullExceptionE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime22LoraCacheFullException22LoraCacheFullExceptionERKNSt6stringE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime22LoraCacheFullException22LoraCacheFullExceptionERKNSt6stringE"></span><span id="_CPPv2N12tensorrt_llm7runtime22LoraCacheFullException22LoraCacheFullExceptionERKNSt6stringE"></span><span id="tensorrt_llm::runtime::LoraCacheFullException::LoraCacheFullException__ssCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCacheFullException_1aefa74dd5535c638a6b5d16d886db0887"></span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">LoraCacheFullException</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">msg</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime22LoraCacheFullException22LoraCacheFullExceptionERKNSt6stringE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime22LoraCacheFullExceptionD0Ev">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime22LoraCacheFullExceptionD0Ev"></span><span id="_CPPv2N12tensorrt_llm7runtime22LoraCacheFullExceptionD0Ev"></span><span id="tensorrt_llm::runtime::LoraCacheFullException::~LoraCacheFullException"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCacheFullException_1a7ab9608608f16e2186d8fedd599b1c26"></span><span class="sig-name descname"><span class="n"><span class="pre">~LoraCacheFullException</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime22LoraCacheFullExceptionD0Ev" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime20LoraCachePageManagerE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime20LoraCachePageManagerE"></span><span id="_CPPv2N12tensorrt_llm7runtime20LoraCachePageManagerE"></span><span id="tensorrt_llm::runtime::LoraCachePageManager"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManager"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">LoraCachePageManager</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime20LoraCachePageManagerE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="docutils container">
|
||
<em>#include <loraCache.h></em></div>
|
||
<p>Holds memory of lora cache pages, and manages allocation and freeing of whole pages. Memory is pre-allocated either on the host or device</p>
|
||
<p>Note that this class is not thread safe </p>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime20LoraCachePageManager9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime20LoraCachePageManager9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManager_1a8ddec7e4c764aefb6ca1e7983f10f6c6"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager9TensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager20LoraCachePageManagerERK26LoraCachePageManagerConfigRK13BufferManager">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime20LoraCachePageManager20LoraCachePageManagerERK26LoraCachePageManagerConfigRK13BufferManager"></span><span id="_CPPv2N12tensorrt_llm7runtime20LoraCachePageManager20LoraCachePageManagerERK26LoraCachePageManagerConfigRK13BufferManager"></span><span id="tensorrt_llm::runtime::LoraCachePageManager::LoraCachePageManager__LoraCachePageManagerConfigCR.BufferManagerCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManager_1a4f03d18ae0cb624f74e470baec8e4cd4"></span><span class="sig-name descname"><span class="n"><span class="pre">LoraCachePageManager</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfigE" title="tensorrt_llm::runtime::LoraCachePageManagerConfig"><span class="n"><span class="pre">LoraCachePageManagerConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">config</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManagerE" title="tensorrt_llm::runtime::BufferManager"><span class="n"><span class="pre">BufferManager</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">bufferManager</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager20LoraCachePageManagerERK26LoraCachePageManagerConfigRK13BufferManager" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><ul class="simple">
|
||
<li><p><strong>config</strong> – <strong>[in]</strong> a <a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig"><span class="std std-ref">LoraCachePageManagerConfig</span></a></p></li>
|
||
<li><p><strong>bufferManager</strong> – <strong>[in]</strong> a Buffermanager used to allocate page blocks </p></li>
|
||
</ul>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager10claimPagesE10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime20LoraCachePageManager10claimPagesE10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime20LoraCachePageManager10claimPagesE10SizeType32"></span><span id="tensorrt_llm::runtime::LoraCachePageManager::claimPages__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManager_1aa931d613427ce9df18abdac708fcec01"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">claimPages</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">numPages</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager10claimPagesE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>claim pages </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>numPages</strong> – <strong>[in]</strong> number of pages to claim </p>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>a tuple, where the first values is a boolean indicating whether pages were claimed. If the first value is true the second value will have a list of pageIds </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime20LoraCachePageManager17numAvailablePagesEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime20LoraCachePageManager17numAvailablePagesEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime20LoraCachePageManager17numAvailablePagesEv"></span><span id="tensorrt_llm::runtime::LoraCachePageManager::numAvailablePagesC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManager_1ab17136bfabc750108a3ab318e60d5d56"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">numAvailablePages</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime20LoraCachePageManager17numAvailablePagesEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>get number of available (free) pages in manager </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>number of free pages in manager </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager12releasePagesERKNSt6vectorINSt6size_tEEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime20LoraCachePageManager12releasePagesERKNSt6vectorINSt6size_tEEE"></span><span id="_CPPv2N12tensorrt_llm7runtime20LoraCachePageManager12releasePagesERKNSt6vectorINSt6size_tEEE"></span><span id="tensorrt_llm::runtime::LoraCachePageManager::releasePages__std::vector:std::s:CR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManager_1afc927205186a82307fc039f95ce2e0e1"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">releasePages</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">pages</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager12releasePagesERKNSt6vectorINSt6size_tEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>release given pages </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>pages</strong> – <strong>[in]</strong> list of pages to release (free) </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime20LoraCachePageManager8blockPtrE10SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime20LoraCachePageManager8blockPtrE10SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime20LoraCachePageManager8blockPtrE10SizeType32"></span><span id="tensorrt_llm::runtime::LoraCachePageManager::blockPtr__SizeType32C"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManager_1a61181364d46c445940a2940e6ca3d469"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor14SharedConstPtrE" title="tensorrt_llm::runtime::ITensor::SharedConstPtr"><span class="n"><span class="pre">SharedConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">blockPtr</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">blockIdx</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime20LoraCachePageManager8blockPtrE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>return pointer to given page block </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>blockIdx;</strong> – <strong>[in]</strong> </p>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>&#8212; pointer to page block </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime20LoraCachePageManager7pagePtrENSt6size_tE">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime20LoraCachePageManager7pagePtrENSt6size_tE"></span><span id="_CPPv2NK12tensorrt_llm7runtime20LoraCachePageManager7pagePtrENSt6size_tE"></span><span id="tensorrt_llm::runtime::LoraCachePageManager::pagePtr__std::sC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManager_1a1a30baf3b114912fccbcdf2ad649755a"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor14SharedConstPtrE" title="tensorrt_llm::runtime::ITensor::SharedConstPtr"><span class="n"><span class="pre">SharedConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">pagePtr</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">pageIdx</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime20LoraCachePageManager7pagePtrENSt6size_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>return pointer to given page </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>pageIdx</strong> – <strong>[in]</strong> </p>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>&#8212; const pointer to page </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager14mutablePagePtrENSt6size_tE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime20LoraCachePageManager14mutablePagePtrENSt6size_tE"></span><span id="_CPPv2N12tensorrt_llm7runtime20LoraCachePageManager14mutablePagePtrENSt6size_tE"></span><span id="tensorrt_llm::runtime::LoraCachePageManager::mutablePagePtr__std::s"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManager_1a8e52eebebbe95fb739988456aded1b2a"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mutablePagePtr</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">pageIdx</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager14mutablePagePtrENSt6size_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>return pointer to given page </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>pageIdx</strong> – <strong>[in]</strong> </p>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>&#8212; mutable pointer to page </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-functions">Private Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager10initializeERK13BufferManager">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime20LoraCachePageManager10initializeERK13BufferManager"></span><span id="_CPPv2N12tensorrt_llm7runtime20LoraCachePageManager10initializeERK13BufferManager"></span><span id="tensorrt_llm::runtime::LoraCachePageManager::initialize__BufferManagerCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManager_1a66be850fa0c5232ea5cbc330bc363b5a"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">initialize</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManagerE" title="tensorrt_llm::runtime::BufferManager"><span class="n"><span class="pre">BufferManager</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">bufferManager</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager10initializeERK13BufferManager" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-members">Private Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager11mPageBlocksE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime20LoraCachePageManager11mPageBlocksE"></span><span id="_CPPv2N12tensorrt_llm7runtime20LoraCachePageManager11mPageBlocksE"></span><span id="tensorrt_llm::runtime::LoraCachePageManager::mPageBlocks__std::vector:TensorPtr:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManager_1af4e201033f3d2b283a526e204d6f9731"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager9TensorPtrE" title="tensorrt_llm::runtime::LoraCachePageManager::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mPageBlocks</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager11mPageBlocksE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager12mFreePageIdsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime20LoraCachePageManager12mFreePageIdsE"></span><span id="_CPPv2N12tensorrt_llm7runtime20LoraCachePageManager12mFreePageIdsE"></span><span id="tensorrt_llm::runtime::LoraCachePageManager::mFreePageIds__std::deque:std::s:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManager_1a57cfac298b65e0a2d6c539d7aebdb85c"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">deque</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mFreePageIds</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager12mFreePageIdsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager11mIsPageFreeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime20LoraCachePageManager11mIsPageFreeE"></span><span id="_CPPv2N12tensorrt_llm7runtime20LoraCachePageManager11mIsPageFreeE"></span><span id="tensorrt_llm::runtime::LoraCachePageManager::mIsPageFree__std::vector:std::uint8_t:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManager_1a8339dcb267e8b1949248dd4567684cd2"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">uint8_t</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mIsPageFree</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager11mIsPageFreeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager7mConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime20LoraCachePageManager7mConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime20LoraCachePageManager7mConfigE"></span><span id="tensorrt_llm::runtime::LoraCachePageManager::mConfig__LoraCachePageManagerConfigC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManager_1af90e90b30a4712ff3dee148f986e6332"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfigE" title="tensorrt_llm::runtime::LoraCachePageManagerConfig"><span class="n"><span class="pre">LoraCachePageManagerConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mConfig</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime20LoraCachePageManager7mConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime21LoraExpectedExceptionE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime21LoraExpectedExceptionE"></span><span id="_CPPv2N12tensorrt_llm7runtime21LoraExpectedExceptionE"></span><span id="tensorrt_llm::runtime::LoraExpectedException"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraExpectedException"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">LoraExpectedException</span></span></span><span class="w"> </span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="k"><span class="pre">public</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">runtime_error</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime21LoraExpectedExceptionE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Subclassed by <a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1LoraCacheFullException"><span class="std std-ref">tensorrt_llm::runtime::LoraCacheFullException</span></a></p>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime21LoraExpectedException21LoraExpectedExceptionERKNSt6stringE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime21LoraExpectedException21LoraExpectedExceptionERKNSt6stringE"></span><span id="_CPPv2N12tensorrt_llm7runtime21LoraExpectedException21LoraExpectedExceptionERKNSt6stringE"></span><span id="tensorrt_llm::runtime::LoraExpectedException::LoraExpectedException__ssCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraExpectedException_1a2e41970c7453e194635c9c7ca583bd99"></span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">LoraExpectedException</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">msg</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime21LoraExpectedException21LoraExpectedExceptionERKNSt6stringE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime21LoraExpectedExceptionD0Ev">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime21LoraExpectedExceptionD0Ev"></span><span id="_CPPv2N12tensorrt_llm7runtime21LoraExpectedExceptionD0Ev"></span><span id="tensorrt_llm::runtime::LoraExpectedException::~LoraExpectedException"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraExpectedException_1a84b390f6dbb652b8501868df858aab64"></span><span class="sig-name descname"><span class="n"><span class="pre">~LoraExpectedException</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime21LoraExpectedExceptionD0Ev" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="loracachepagemanagerconfig-h">
|
||
<h2>loraCachePageManagerConfig.h<a class="headerlink" href="#loracachepagemanagerconfig-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-functions">Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtimelsERNSt7ostreamERK26LoraCachePageManagerConfig">
|
||
<span id="_CPPv3N12tensorrt_llm7runtimelsERNSt7ostreamERK26LoraCachePageManagerConfig"></span><span id="_CPPv2N12tensorrt_llm7runtimelsERNSt7ostreamERK26LoraCachePageManagerConfig"></span><span id="tensorrt_llm::runtime::lshift-operator__osR.LoraCachePageManagerConfigCR"></span><span class="target" id="loraCachePageManagerConfig_8h_1a02fbd0e78a87e57947a3d24b50f5141a"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ostream</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="k"><span class="pre">operator</span></span><span class="o"><span class="pre"><<</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ostream</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">os</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfigE" title="tensorrt_llm::runtime::LoraCachePageManagerConfig"><span class="n"><span class="pre">LoraCachePageManagerConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">c</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtimelsERNSt7ostreamERK26LoraCachePageManagerConfig" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9to_stringERK26LoraCachePageManagerConfig">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9to_stringERK26LoraCachePageManagerConfig"></span><span id="_CPPv2N12tensorrt_llm7runtime9to_stringERK26LoraCachePageManagerConfig"></span><span id="tensorrt_llm::runtime::to_string__LoraCachePageManagerConfigCR"></span><span class="target" id="loraCachePageManagerConfig_8h_1ab560cb107debf46c244846e86a6af35a"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">to_string</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfigE" title="tensorrt_llm::runtime::LoraCachePageManagerConfig"><span class="n"><span class="pre">LoraCachePageManagerConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">c</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9to_stringERK26LoraCachePageManagerConfig" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26LoraCachePageManagerConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime26LoraCachePageManagerConfigE"></span><span id="tensorrt_llm::runtime::LoraCachePageManagerConfig"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">LoraCachePageManagerConfig</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="docutils container">
|
||
<em>#include <loraCachePageManagerConfig.h></em></div>
|
||
<p>Configuration for <a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1LoraCachePageManager"><span class="std std-ref">LoraCachePageManager</span></a></p>
|
||
<p>See <a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1LoraCache"><span class="std std-ref">LoraCache</span></a> docs for description of pages, slots, and page blocks. </p>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig26LoraCachePageManagerConfigEN7runtime10MemoryTypeEN8nvinfer18DataTypeE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26LoraCachePageManagerConfig26LoraCachePageManagerConfigEN7runtime10MemoryTypeEN8nvinfer18DataTypeE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime26LoraCachePageManagerConfig26LoraCachePageManagerConfigEN7runtime10MemoryTypeEN8nvinfer18DataTypeE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType32"></span><span id="tensorrt_llm::runtime::LoraCachePageManagerConfig::LoraCachePageManagerConfig__runtime::MemoryType.nvinfer1::DataType.SizeType32.SizeType32.SizeType32.SizeType32.SizeType32CE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig_1a6e1aff651ea70a7fb83e92ee5b6973bf"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">LoraCachePageManagerConfig</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="tensorrt_llm::runtime::MemoryType"><span class="n"><span class="pre">MemoryType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">memType</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">dType</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">totalNumPages</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxPagesPerBlock</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">slotsPerPage</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">pageWidth</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">numCopyStreams</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig26LoraCachePageManagerConfigEN7runtime10MemoryTypeEN8nvinfer18DataTypeE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime26LoraCachePageManagerConfig13getMemoryTypeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime26LoraCachePageManagerConfig13getMemoryTypeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime26LoraCachePageManagerConfig13getMemoryTypeEv"></span><span id="tensorrt_llm::runtime::LoraCachePageManagerConfig::getMemoryTypeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig_1a1d555270b7def499025cefa19b7abd03"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="tensorrt_llm::runtime::MemoryType"><span class="n"><span class="pre">MemoryType</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getMemoryType</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime26LoraCachePageManagerConfig13getMemoryTypeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig13setMemoryTypeERKN7runtime10MemoryTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26LoraCachePageManagerConfig13setMemoryTypeERKN7runtime10MemoryTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime26LoraCachePageManagerConfig13setMemoryTypeERKN7runtime10MemoryTypeE"></span><span id="tensorrt_llm::runtime::LoraCachePageManagerConfig::setMemoryType__runtime::MemoryTypeCRCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig_1a15aa48ad133fb36443123a0e38796173"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setMemoryType</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="tensorrt_llm::runtime::MemoryType"><span class="n"><span class="pre">MemoryType</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">memoryType</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig13setMemoryTypeERKN7runtime10MemoryTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime26LoraCachePageManagerConfig11getDataTypeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime26LoraCachePageManagerConfig11getDataTypeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime26LoraCachePageManagerConfig11getDataTypeEv"></span><span id="tensorrt_llm::runtime::LoraCachePageManagerConfig::getDataTypeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig_1a5b8569adff45e7ba8fae9147bf42a4a9"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getDataType</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime26LoraCachePageManagerConfig11getDataTypeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig11setDataTypeERKN8nvinfer18DataTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26LoraCachePageManagerConfig11setDataTypeERKN8nvinfer18DataTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime26LoraCachePageManagerConfig11setDataTypeERKN8nvinfer18DataTypeE"></span><span id="tensorrt_llm::runtime::LoraCachePageManagerConfig::setDataType__nvinfer1::DataTypeCRCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig_1ac6c9e32d3645e715c9114aabb0a3aed1"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setDataType</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">dtype</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig11setDataTypeERKN8nvinfer18DataTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime26LoraCachePageManagerConfig16getTotalNumPagesEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime26LoraCachePageManagerConfig16getTotalNumPagesEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime26LoraCachePageManagerConfig16getTotalNumPagesEv"></span><span id="tensorrt_llm::runtime::LoraCachePageManagerConfig::getTotalNumPagesCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig_1ac78c314b429c387bf16594e813aec5fd"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getTotalNumPages</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime26LoraCachePageManagerConfig16getTotalNumPagesEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig15setTotalNumPageERK10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26LoraCachePageManagerConfig15setTotalNumPageERK10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime26LoraCachePageManagerConfig15setTotalNumPageERK10SizeType32"></span><span id="tensorrt_llm::runtime::LoraCachePageManagerConfig::setTotalNumPage__SizeType32CRCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig_1afb9b7258d161eb6069a1cf568ac3e2f3"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setTotalNumPage</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">totalNumPages</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig15setTotalNumPageERK10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime26LoraCachePageManagerConfig19getMaxPagesPerBlockEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime26LoraCachePageManagerConfig19getMaxPagesPerBlockEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime26LoraCachePageManagerConfig19getMaxPagesPerBlockEv"></span><span id="tensorrt_llm::runtime::LoraCachePageManagerConfig::getMaxPagesPerBlockCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig_1abcbd157aba25cb7c6b71961077b3ac4a"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getMaxPagesPerBlock</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime26LoraCachePageManagerConfig19getMaxPagesPerBlockEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig19setMaxPagesPerBlockERK10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26LoraCachePageManagerConfig19setMaxPagesPerBlockERK10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime26LoraCachePageManagerConfig19setMaxPagesPerBlockERK10SizeType32"></span><span id="tensorrt_llm::runtime::LoraCachePageManagerConfig::setMaxPagesPerBlock__SizeType32CRCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig_1aa3f4219a0ddab455d727fe54fc834999"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setMaxPagesPerBlock</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">maxPagesPerBlock</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig19setMaxPagesPerBlockERK10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime26LoraCachePageManagerConfig15getSlotsPerPageEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime26LoraCachePageManagerConfig15getSlotsPerPageEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime26LoraCachePageManagerConfig15getSlotsPerPageEv"></span><span id="tensorrt_llm::runtime::LoraCachePageManagerConfig::getSlotsPerPageCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig_1a0cb73d5d68b2ff2aa05e55c65aabb6eb"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getSlotsPerPage</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime26LoraCachePageManagerConfig15getSlotsPerPageEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig15setSlotsPerPageERK10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26LoraCachePageManagerConfig15setSlotsPerPageERK10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime26LoraCachePageManagerConfig15setSlotsPerPageERK10SizeType32"></span><span id="tensorrt_llm::runtime::LoraCachePageManagerConfig::setSlotsPerPage__SizeType32CRCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig_1a2741800c4c8b6b550af4894410acfecb"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setSlotsPerPage</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">slotsPerPage</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig15setSlotsPerPageERK10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime26LoraCachePageManagerConfig12getPageWidthEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime26LoraCachePageManagerConfig12getPageWidthEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime26LoraCachePageManagerConfig12getPageWidthEv"></span><span id="tensorrt_llm::runtime::LoraCachePageManagerConfig::getPageWidthCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig_1add88d63cdf4bd8a6220a2662709b6e4b"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getPageWidth</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime26LoraCachePageManagerConfig12getPageWidthEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig12setPageWidthERK10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26LoraCachePageManagerConfig12setPageWidthERK10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime26LoraCachePageManagerConfig12setPageWidthERK10SizeType32"></span><span id="tensorrt_llm::runtime::LoraCachePageManagerConfig::setPageWidth__SizeType32CRCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig_1ac79621f57044a1cb648067d9e0594347"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setPageWidth</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">pageWidth</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig12setPageWidthERK10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime26LoraCachePageManagerConfig13getInitToZeroEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime26LoraCachePageManagerConfig13getInitToZeroEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime26LoraCachePageManagerConfig13getInitToZeroEv"></span><span id="tensorrt_llm::runtime::LoraCachePageManagerConfig::getInitToZeroCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig_1a4550e479df29f722b2c75e8b7614de45"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getInitToZero</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime26LoraCachePageManagerConfig13getInitToZeroEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig13setInitToZeroEb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26LoraCachePageManagerConfig13setInitToZeroEb"></span><span id="_CPPv2N12tensorrt_llm7runtime26LoraCachePageManagerConfig13setInitToZeroEb"></span><span id="tensorrt_llm::runtime::LoraCachePageManagerConfig::setInitToZero__bCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig_1a7f940837905e42adec2d6ed678a3677f"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setInitToZero</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">initToZero</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig13setInitToZeroEb" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime26LoraCachePageManagerConfig17getNumCopyStreamsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime26LoraCachePageManagerConfig17getNumCopyStreamsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime26LoraCachePageManagerConfig17getNumCopyStreamsEv"></span><span id="tensorrt_llm::runtime::LoraCachePageManagerConfig::getNumCopyStreamsCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig_1a3928d560911fb6b831a1ab11b5ae60dc"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getNumCopyStreams</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime26LoraCachePageManagerConfig17getNumCopyStreamsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig17setNumCopyStreamsE10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26LoraCachePageManagerConfig17setNumCopyStreamsE10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime26LoraCachePageManagerConfig17setNumCopyStreamsE10SizeType32"></span><span id="tensorrt_llm::runtime::LoraCachePageManagerConfig::setNumCopyStreams__SizeType32CE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig_1a0a8555fa27f30b63a6d13fd4698303d6"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setNumCopyStreams</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">numCopyStreams</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig17setNumCopyStreamsE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-members">Private Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig11mMemoryTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26LoraCachePageManagerConfig11mMemoryTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime26LoraCachePageManagerConfig11mMemoryTypeE"></span><span id="tensorrt_llm::runtime::LoraCachePageManagerConfig::mMemoryType__runtime::MemoryType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig_1a20d70126704e1f716dfcaf06da538b86"></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="tensorrt_llm::runtime::MemoryType"><span class="n"><span class="pre">MemoryType</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mMemoryType</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig11mMemoryTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig9mDataTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26LoraCachePageManagerConfig9mDataTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime26LoraCachePageManagerConfig9mDataTypeE"></span><span id="tensorrt_llm::runtime::LoraCachePageManagerConfig::mDataType__nvinfer1::DataType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig_1a016b0e9c25c6ee1587448dee9059b809"></span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mDataType</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig9mDataTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig14mTotalNumPagesE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26LoraCachePageManagerConfig14mTotalNumPagesE"></span><span id="_CPPv2N12tensorrt_llm7runtime26LoraCachePageManagerConfig14mTotalNumPagesE"></span><span id="tensorrt_llm::runtime::LoraCachePageManagerConfig::mTotalNumPages__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig_1a1517a8ee5accef3c6d7b88dbb6766ce0"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mTotalNumPages</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig14mTotalNumPagesE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig17mMaxPagesPerBlockE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26LoraCachePageManagerConfig17mMaxPagesPerBlockE"></span><span id="_CPPv2N12tensorrt_llm7runtime26LoraCachePageManagerConfig17mMaxPagesPerBlockE"></span><span id="tensorrt_llm::runtime::LoraCachePageManagerConfig::mMaxPagesPerBlock__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig_1a0f9187333756785772166e1302f3c4da"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mMaxPagesPerBlock</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig17mMaxPagesPerBlockE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig13mSlotsPerPageE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26LoraCachePageManagerConfig13mSlotsPerPageE"></span><span id="_CPPv2N12tensorrt_llm7runtime26LoraCachePageManagerConfig13mSlotsPerPageE"></span><span id="tensorrt_llm::runtime::LoraCachePageManagerConfig::mSlotsPerPage__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig_1a4990fcbf90a213977985fbd41ac2f22c"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mSlotsPerPage</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig13mSlotsPerPageE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig10mPageWidthE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26LoraCachePageManagerConfig10mPageWidthE"></span><span id="_CPPv2N12tensorrt_llm7runtime26LoraCachePageManagerConfig10mPageWidthE"></span><span id="tensorrt_llm::runtime::LoraCachePageManagerConfig::mPageWidth__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig_1a57b18ff225e47d5e1f41c9962c85cad9"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mPageWidth</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig10mPageWidthE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig15mNumCopyStreamsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26LoraCachePageManagerConfig15mNumCopyStreamsE"></span><span id="_CPPv2N12tensorrt_llm7runtime26LoraCachePageManagerConfig15mNumCopyStreamsE"></span><span id="tensorrt_llm::runtime::LoraCachePageManagerConfig::mNumCopyStreams__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig_1a5abcc91f11d68bcbb50ba133ece478b6"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mNumCopyStreams</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">1</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig15mNumCopyStreamsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig11mInitToZeroE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26LoraCachePageManagerConfig11mInitToZeroE"></span><span id="_CPPv2N12tensorrt_llm7runtime26LoraCachePageManagerConfig11mInitToZeroE"></span><span id="tensorrt_llm::runtime::LoraCachePageManagerConfig::mInitToZero__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraCachePageManagerConfig_1a010768dca591da9134e770156e44a141"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mInitToZero</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26LoraCachePageManagerConfig11mInitToZeroE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="loramodule-h">
|
||
<h2>loraModule.h<a class="headerlink" href="#loramodule-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-functions">Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtimelsERNSt7ostreamERK10LoraModule">
|
||
<span id="_CPPv3N12tensorrt_llm7runtimelsERNSt7ostreamERK10LoraModule"></span><span id="_CPPv2N12tensorrt_llm7runtimelsERNSt7ostreamERK10LoraModule"></span><span id="tensorrt_llm::runtime::lshift-operator__osR.LoraModuleCR"></span><span class="target" id="loraModule_8h_1a019fd70ba84e9b865bc6b7b58db3fd6f"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ostream</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="k"><span class="pre">operator</span></span><span class="o"><span class="pre"><<</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ostream</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">output</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModuleE" title="tensorrt_llm::runtime::LoraModule"><span class="n"><span class="pre">LoraModule</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">module</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtimelsERNSt7ostreamERK10LoraModule" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModuleE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModuleE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModuleE"></span><span id="tensorrt_llm::runtime::LoraModule"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">LoraModule</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModuleE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp enum-class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule10ModuleTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule10ModuleTypeE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a784821fbc7f44239dcf33e75bad249b8"></span><span class="k"><span class="pre">enum</span></span><span class="w"> </span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">ModuleType</span></span></span><span class="w"> </span><span class="p"><span class="pre">:</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p><em>Values:</em></p>
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType8kINVALIDE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule10ModuleType8kINVALIDE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule10ModuleType8kINVALIDE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a784821fbc7f44239dcf33e75bad249b8a90e668d1015c46c274bdcf183d2bfd30"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kINVALID</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType8kINVALIDE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType9kATTN_QKVE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule10ModuleType9kATTN_QKVE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule10ModuleType9kATTN_QKVE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a784821fbc7f44239dcf33e75bad249b8a3abaa6d49d41464821a96fe45ae499f3"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kATTN_QKV</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType9kATTN_QKVE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType7kATTN_QE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule10ModuleType7kATTN_QE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule10ModuleType7kATTN_QE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a784821fbc7f44239dcf33e75bad249b8a885b263f8c935fc2e9a227e7fb7bfc85"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kATTN_Q</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType7kATTN_QE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType7kATTN_KE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule10ModuleType7kATTN_KE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule10ModuleType7kATTN_KE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a784821fbc7f44239dcf33e75bad249b8a1d15211064de5ce58d9aa8b3f54130f2"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kATTN_K</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType7kATTN_KE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType7kATTN_VE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule10ModuleType7kATTN_VE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule10ModuleType7kATTN_VE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a784821fbc7f44239dcf33e75bad249b8aed04252974b4c95c0691bbbd7d554557"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kATTN_V</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType7kATTN_VE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType11kATTN_DENSEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule10ModuleType11kATTN_DENSEE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule10ModuleType11kATTN_DENSEE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a784821fbc7f44239dcf33e75bad249b8a9678d862c1ecefc4790acc3076a9ae39"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kATTN_DENSE</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType11kATTN_DENSEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType12kMLP_H_TO_4HE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule10ModuleType12kMLP_H_TO_4HE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule10ModuleType12kMLP_H_TO_4HE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a784821fbc7f44239dcf33e75bad249b8a5897d380b701b8c819f6452783ea76a6"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kMLP_H_TO_4H</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType12kMLP_H_TO_4HE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType12kMLP_4H_TO_HE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule10ModuleType12kMLP_4H_TO_HE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule10ModuleType12kMLP_4H_TO_HE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a784821fbc7f44239dcf33e75bad249b8a1291a3a4a4726a507e30b07cecc754b8"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kMLP_4H_TO_H</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType12kMLP_4H_TO_HE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType9kMLP_GATEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule10ModuleType9kMLP_GATEE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule10ModuleType9kMLP_GATEE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a784821fbc7f44239dcf33e75bad249b8a5414fc337719904c181fb53f649a6a02"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kMLP_GATE</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType9kMLP_GATEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType15kCROSS_ATTN_QKVE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule10ModuleType15kCROSS_ATTN_QKVE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule10ModuleType15kCROSS_ATTN_QKVE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a784821fbc7f44239dcf33e75bad249b8a7b69854e31937cdceb210086feefea65"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kCROSS_ATTN_QKV</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType15kCROSS_ATTN_QKVE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType13kCROSS_ATTN_QE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule10ModuleType13kCROSS_ATTN_QE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule10ModuleType13kCROSS_ATTN_QE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a784821fbc7f44239dcf33e75bad249b8a82f7bdb5649b5d0f8a90fa1ebb5c1839"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kCROSS_ATTN_Q</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType13kCROSS_ATTN_QE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType13kCROSS_ATTN_KE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule10ModuleType13kCROSS_ATTN_KE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule10ModuleType13kCROSS_ATTN_KE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a784821fbc7f44239dcf33e75bad249b8a6a2b7e8676549d9de16f408efc6a2614"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kCROSS_ATTN_K</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType13kCROSS_ATTN_KE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType13kCROSS_ATTN_VE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule10ModuleType13kCROSS_ATTN_VE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule10ModuleType13kCROSS_ATTN_VE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a784821fbc7f44239dcf33e75bad249b8a0b7fee7135c586e02ea30c3b61e59c7c"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kCROSS_ATTN_V</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType13kCROSS_ATTN_VE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType17kCROSS_ATTN_DENSEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule10ModuleType17kCROSS_ATTN_DENSEE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule10ModuleType17kCROSS_ATTN_DENSEE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a784821fbc7f44239dcf33e75bad249b8a4831dc4fe8b758988c399d8b3368fcd4"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kCROSS_ATTN_DENSE</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType17kCROSS_ATTN_DENSEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType12kMOE_H_TO_4HE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule10ModuleType12kMOE_H_TO_4HE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule10ModuleType12kMOE_H_TO_4HE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a784821fbc7f44239dcf33e75bad249b8a3f915fd485fd5c23d418effadcdef87c"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kMOE_H_TO_4H</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType12kMOE_H_TO_4HE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType12kMOE_4H_TO_HE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule10ModuleType12kMOE_4H_TO_HE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule10ModuleType12kMOE_4H_TO_HE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a784821fbc7f44239dcf33e75bad249b8a16cb51bf6af53b8657c999409aa1c5e8"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kMOE_4H_TO_H</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType12kMOE_4H_TO_HE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType9kMOE_GATEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule10ModuleType9kMOE_GATEE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule10ModuleType9kMOE_GATEE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a784821fbc7f44239dcf33e75bad249b8a853eac337a8fbded3c1ffffe156cbc8e"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kMOE_GATE</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType9kMOE_GATEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType11kMOE_ROUTERE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule10ModuleType11kMOE_ROUTERE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule10ModuleType11kMOE_ROUTERE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a784821fbc7f44239dcf33e75bad249b8a743f330fc64cb3653e638bdd503ad2b3"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kMOE_ROUTER</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType11kMOE_ROUTERE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType11kMLP_ROUTERE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule10ModuleType11kMLP_ROUTERE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule10ModuleType11kMLP_ROUTERE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a784821fbc7f44239dcf33e75bad249b8a21e30455fd5abaff942fb59aefd875c4"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kMLP_ROUTER</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleType11kMLP_ROUTERE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a894d3a0fd7f335adb87a0f4d8efe672a"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule9TensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule10LoraModuleERK10ModuleType10SizeType3210SizeType32bb10SizeType3210SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule10LoraModuleERK10ModuleType10SizeType3210SizeType32bb10SizeType3210SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule10LoraModuleERK10ModuleType10SizeType3210SizeType32bb10SizeType3210SizeType32"></span><span id="tensorrt_llm::runtime::LoraModule::LoraModule__ModuleTypeCR.SizeType32.SizeType32.b.b.SizeType32.SizeType32CE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1acb1372f4c7079235fb8e594691919fee"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">LoraModule</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleTypeE" title="tensorrt_llm::runtime::LoraModule::ModuleType"><span class="n"><span class="pre">ModuleType</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">t</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">inDim</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">outDim</span></span>, <span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">inDimFirst</span></span>, <span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">outDimFirst</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">inTpSplitDim</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">outTpSplitDim</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10LoraModuleERK10ModuleType10SizeType3210SizeType32bb10SizeType3210SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule10LoraModuleEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule10LoraModuleEv"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule10LoraModuleEv"></span><span id="tensorrt_llm::runtime::LoraModule::LoraModuleCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a010f32ba204263e4ce58494e02ea9198"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">LoraModule</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10LoraModuleEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule10LoraModuleERK10LoraModule">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule10LoraModuleERK10LoraModule"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule10LoraModuleERK10LoraModule"></span><span id="tensorrt_llm::runtime::LoraModule::LoraModule__LoraModuleCRCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1ade6c5f359962e48848102880cfa72fa2"></span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">LoraModule</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10LoraModuleERK10LoraModule" title="tensorrt_llm::runtime::LoraModule::LoraModule"><span class="n"><span class="pre">LoraModule</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">o</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">default</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10LoraModuleERK10LoraModule" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModuleaSERK10LoraModule">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModuleaSERK10LoraModule"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModuleaSERK10LoraModule"></span><span id="tensorrt_llm::runtime::LoraModule::assign-operator__LoraModuleCRCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a0318026e4d60431ed806daa1807dcc23"></span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModuleE" title="tensorrt_llm::runtime::LoraModule"><span class="n"><span class="pre">LoraModule</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="k"><span class="pre">operator</span></span><span class="o"><span class="pre">=</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModuleE" title="tensorrt_llm::runtime::LoraModule"><span class="n"><span class="pre">LoraModule</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">o</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">default</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModuleaSERK10LoraModule" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10LoraModule18flattenedInOutSizeE10SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10LoraModule18flattenedInOutSizeE10SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime10LoraModule18flattenedInOutSizeE10SizeType32"></span><span id="tensorrt_llm::runtime::LoraModule::flattenedInOutSize__SizeType32CCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a1994de44e740f5f3df9f70e2bae93093"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">flattenedInOutSize</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">adapterSize</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule18flattenedInOutSizeE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10LoraModule6inSizeE10SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10LoraModule6inSizeE10SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime10LoraModule6inSizeE10SizeType32"></span><span id="tensorrt_llm::runtime::LoraModule::inSize__SizeType32CCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a4cbb632fe6af3a63048df1392aaabaa3"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">inSize</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">adapterSize</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule6inSizeE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10LoraModule7outSizeE10SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10LoraModule7outSizeE10SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime10LoraModule7outSizeE10SizeType32"></span><span id="tensorrt_llm::runtime::LoraModule::outSize__SizeType32CCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1ad7cf8314d5f8fb50d8dc4c10ea778674"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">outSize</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">adapterSize</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule7outSizeE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10LoraModule11localInSizeE10SizeType3210SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10LoraModule11localInSizeE10SizeType3210SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime10LoraModule11localInSizeE10SizeType3210SizeType32"></span><span id="tensorrt_llm::runtime::LoraModule::localInSize__SizeType32.SizeType32CCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a59babff6e5dea96b528bc6be8e42a28f"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">localInSize</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">adapterSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tpSize</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule11localInSizeE10SizeType3210SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10LoraModule12localOutSizeE10SizeType3210SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10LoraModule12localOutSizeE10SizeType3210SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime10LoraModule12localOutSizeE10SizeType3210SizeType32"></span><span id="tensorrt_llm::runtime::LoraModule::localOutSize__SizeType32.SizeType32CCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a3e6970a72ca8287ce0c4baac39cf9497"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">localOutSize</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">adapterSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tpSize</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule12localOutSizeE10SizeType3210SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10LoraModule10localInDimE10SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10LoraModule10localInDimE10SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime10LoraModule10localInDimE10SizeType32"></span><span id="tensorrt_llm::runtime::LoraModule::localInDim__SizeType32CCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1ac86684c323e7e21d1b46964ce68580e4"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">localInDim</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tpSize</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule10localInDimE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10LoraModule11localOutDimE10SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10LoraModule11localOutDimE10SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime10LoraModule11localOutDimE10SizeType32"></span><span id="tensorrt_llm::runtime::LoraModule::localOutDim__SizeType32CCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a7f5d2a328ee5e0b45e82053e13016168"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">localOutDim</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tpSize</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule11localOutDimE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10LoraModule18localInAdapterSizeE10SizeType3210SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10LoraModule18localInAdapterSizeE10SizeType3210SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime10LoraModule18localInAdapterSizeE10SizeType3210SizeType32"></span><span id="tensorrt_llm::runtime::LoraModule::localInAdapterSize__SizeType32.SizeType32CCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a71b480241e25d9d3a6b33c444f790bff"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">localInAdapterSize</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">adapterSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tpSize</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule18localInAdapterSizeE10SizeType3210SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10LoraModule19localOutAdapterSizeE10SizeType3210SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10LoraModule19localOutAdapterSizeE10SizeType3210SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime10LoraModule19localOutAdapterSizeE10SizeType3210SizeType32"></span><span id="tensorrt_llm::runtime::LoraModule::localOutAdapterSize__SizeType32.SizeType32CCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1aa8d74dff461cca4c35017e85bafd6078"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">localOutAdapterSize</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">adapterSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tpSize</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule19localOutAdapterSizeE10SizeType3210SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10LoraModule14localInOutSizeE10SizeType3210SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10LoraModule14localInOutSizeE10SizeType3210SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime10LoraModule14localInOutSizeE10SizeType3210SizeType32"></span><span id="tensorrt_llm::runtime::LoraModule::localInOutSize__SizeType32.SizeType32CCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a6884e16ab6f2d35329f34a8194870496"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">localInOutSize</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">adapterSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tpSize</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule14localInOutSizeE10SizeType3210SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10LoraModule5valueEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10LoraModule5valueEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime10LoraModule5valueEv"></span><span id="tensorrt_llm::runtime::LoraModule::valueCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1ac3fa242daa4252b826531467fda176f2"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule5valueEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10LoraModule4nameEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10LoraModule4nameEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime10LoraModule4nameEv"></span><span id="tensorrt_llm::runtime::LoraModule::nameCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a6d3d51b4414933c97d3e5a64d08e7d0f"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string_view</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">name</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule4nameEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10LoraModule5inDimEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10LoraModule5inDimEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime10LoraModule5inDimEv"></span><span id="tensorrt_llm::runtime::LoraModule::inDimCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a221a2ce81aafff9feae5fd3a7785b3a4"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">inDim</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule5inDimEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10LoraModule6outDimEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10LoraModule6outDimEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime10LoraModule6outDimEv"></span><span id="tensorrt_llm::runtime::LoraModule::outDimCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a638a5cfeb52c2343106ba7e700a2abd6"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">outDim</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule6outDimEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10LoraModule10inDimFirstEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10LoraModule10inDimFirstEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime10LoraModule10inDimFirstEv"></span><span id="tensorrt_llm::runtime::LoraModule::inDimFirstCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1ac380116029394b465e212009b2151e16"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">inDimFirst</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule10inDimFirstEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10LoraModule11outDimFirstEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10LoraModule11outDimFirstEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime10LoraModule11outDimFirstEv"></span><span id="tensorrt_llm::runtime::LoraModule::outDimFirstCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a30b637769120876b91d81f93fcf4fcc1"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">outDimFirst</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule11outDimFirstEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10LoraModule12inTpSplitDimEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10LoraModule12inTpSplitDimEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime10LoraModule12inTpSplitDimEv"></span><span id="tensorrt_llm::runtime::LoraModule::inTpSplitDimCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1ad81da0b83ac2896ea83bef6edbe5c301"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">inTpSplitDim</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule12inTpSplitDimEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10LoraModule13outTpSplitDimEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10LoraModule13outTpSplitDimEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime10LoraModule13outTpSplitDimEv"></span><span id="tensorrt_llm::runtime::LoraModule::outTpSplitDimCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1aee2e5e7995d7445e61e99466a0671c6c"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">outTpSplitDim</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10LoraModule13outTpSplitDimEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-functions">Public Static Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule17createLoraModulesERKNSt6vectorINSt6stringEEE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType3210SizeType3210SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule17createLoraModulesERKNSt6vectorINSt6stringEEE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType3210SizeType3210SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule17createLoraModulesERKNSt6vectorINSt6stringEEE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType3210SizeType3210SizeType32"></span><span id="tensorrt_llm::runtime::LoraModule::createLoraModules__std::vector:ss:CR.SizeType32.SizeType32.SizeType32.SizeType32.SizeType32.SizeType32.SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a93b14a632d0f904290d2a7e9d0debf46"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModuleE" title="tensorrt_llm::runtime::LoraModule"><span class="n"><span class="pre">LoraModule</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">createLoraModules</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">loraModuleNames</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">hiddenSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">mlpHiddenSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">numAttentionHeads</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">numKvAttentionHeads</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">attentionHeadSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tpSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">numExperts</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule17createLoraModulesERKNSt6vectorINSt6stringEEE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType3210SizeType3210SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule12toModuleTypeERKNSt11string_viewE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule12toModuleTypeERKNSt11string_viewE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule12toModuleTypeERKNSt11string_viewE"></span><span id="tensorrt_llm::runtime::LoraModule::toModuleType__std::string_viewCRCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1ab8265ca2adf436d70b3ca2309fe66c96"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleTypeE" title="tensorrt_llm::runtime::LoraModule::ModuleType"><span class="n"><span class="pre">ModuleType</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">toModuleType</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string_view</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">name</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule12toModuleTypeERKNSt11string_viewE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule12toModuleNameE10ModuleType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule12toModuleNameE10ModuleType"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule12toModuleNameE10ModuleType"></span><span id="tensorrt_llm::runtime::LoraModule::toModuleName__ModuleTypeCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a4b2ba83d368ecf7cc04a9554552939a3"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string_view</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">toModuleName</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleTypeE" title="tensorrt_llm::runtime::LoraModule::ModuleType"><span class="n"><span class="pre">ModuleType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">t</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule12toModuleNameE10ModuleType" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule12toModuleNameE10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule12toModuleNameE10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule12toModuleNameE10SizeType32"></span><span id="tensorrt_llm::runtime::LoraModule::toModuleName__SizeType32CE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a3767755d14fbb821ddbbb2d353c54d33"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string_view</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">toModuleName</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">id</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule12toModuleNameE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-members">Private Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule5mTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule5mTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule5mTypeE"></span><span id="tensorrt_llm::runtime::LoraModule::mType__ModuleType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1ae7b52971db55bf920e68ce2d246767e8"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule10ModuleTypeE" title="tensorrt_llm::runtime::LoraModule::ModuleType"><span class="n"><span class="pre">ModuleType</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mType</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule5mTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule6mInDimE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule6mInDimE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule6mInDimE"></span><span id="tensorrt_llm::runtime::LoraModule::mInDim__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a7bb2cb44338e43391db5183d89592547"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mInDim</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule6mInDimE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule7mOutDimE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule7mOutDimE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule7mOutDimE"></span><span id="tensorrt_llm::runtime::LoraModule::mOutDim__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a87596bc34d693acff958dddc44d45f49"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mOutDim</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule7mOutDimE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule11mInDimFirstE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule11mInDimFirstE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule11mInDimFirstE"></span><span id="tensorrt_llm::runtime::LoraModule::mInDimFirst__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1aee316e1b2b3fb3cc31c1f94c6dcea07f"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mInDimFirst</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule11mInDimFirstE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule12mOutDimFirstE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule12mOutDimFirstE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule12mOutDimFirstE"></span><span id="tensorrt_llm::runtime::LoraModule::mOutDimFirst__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1aa04960b7f5fac8ae65016db25dd8f64c"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mOutDimFirst</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule12mOutDimFirstE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule13mInTpSplitDimE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule13mInTpSplitDimE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule13mInTpSplitDimE"></span><span id="tensorrt_llm::runtime::LoraModule::mInTpSplitDim__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a85e27922cbd350fc4b21358d4f43b389"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mInTpSplitDim</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule13mInTpSplitDimE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10LoraModule14mOutTpSplitDimE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10LoraModule14mOutTpSplitDimE"></span><span id="_CPPv2N12tensorrt_llm7runtime10LoraModule14mOutTpSplitDimE"></span><span id="tensorrt_llm::runtime::LoraModule::mOutTpSplitDim__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LoraModule_1a82af046608a063edd7aeda7898de377f"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mOutTpSplitDim</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10LoraModule14mOutTpSplitDimE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="medusamodule-h">
|
||
<h2>medusaModule.h<a class="headerlink" href="#medusamodule-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12MedusaModuleE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12MedusaModuleE"></span><span id="_CPPv2N12tensorrt_llm7runtime12MedusaModuleE"></span><span id="tensorrt_llm::runtime::MedusaModule"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MedusaModule"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">MedusaModule</span></span></span><span class="w"> </span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="k"><span class="pre">public</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModuleE" title="tensorrt_llm::runtime::SpeculativeDecodingModule"><span class="n"><span class="pre">SpeculativeDecodingModule</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12MedusaModuleE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12MedusaModule9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12MedusaModule9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime12MedusaModule9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MedusaModule_1a4ad80835c2e8efd62e459d3cacbc252a"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12MedusaModule9TensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12MedusaModule13MedusaChoicesE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12MedusaModule13MedusaChoicesE"></span><span id="_CPPv2N12tensorrt_llm7runtime12MedusaModule13MedusaChoicesE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MedusaModule_1a811960804e1abe84290dc2d9ef1048df"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">MedusaChoices</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12MedusaModule13MedusaChoicesE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12MedusaModule12MedusaModuleE10SizeType3210SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12MedusaModule12MedusaModuleE10SizeType3210SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime12MedusaModule12MedusaModuleE10SizeType3210SizeType32"></span><span id="tensorrt_llm::runtime::MedusaModule::MedusaModule__SizeType32.SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MedusaModule_1a90e71c121e29000fcedc0a4f69b1191e"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">MedusaModule</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxAcceptedTokens</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxDraftTokens</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12MedusaModule12MedusaModuleE10SizeType3210SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12MedusaModule12MedusaModuleEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12MedusaModule12MedusaModuleEv"></span><span id="_CPPv2N12tensorrt_llm7runtime12MedusaModule12MedusaModuleEv"></span><span id="tensorrt_llm::runtime::MedusaModule::MedusaModule"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MedusaModule_1acd250003687dc783ce61c6d5e03f2da9"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">MedusaModule</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12MedusaModule12MedusaModuleEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime12MedusaModule16getMedusaChoicesEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime12MedusaModule16getMedusaChoicesEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime12MedusaModule16getMedusaChoicesEv"></span><span id="tensorrt_llm::runtime::MedusaModule::getMedusaChoicesC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MedusaModule_1a59c59e09ca553ababc8b4088505160f7"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12MedusaModule13MedusaChoicesE" title="tensorrt_llm::runtime::MedusaModule::MedusaChoices"><span class="n"><span class="pre">MedusaChoices</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="n"><span class="pre">getMedusaChoices</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime12MedusaModule16getMedusaChoicesEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-members">Private Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12MedusaModule21mDefaultMedusaChoicesE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12MedusaModule21mDefaultMedusaChoicesE"></span><span id="_CPPv2N12tensorrt_llm7runtime12MedusaModule21mDefaultMedusaChoicesE"></span><span id="tensorrt_llm::runtime::MedusaModule::mDefaultMedusaChoices__MedusaChoices"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MedusaModule_1aea2c515147409dd184931a4a1fa9ace1"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12MedusaModule13MedusaChoicesE" title="tensorrt_llm::runtime::MedusaModule::MedusaChoices"><span class="n"><span class="pre">MedusaChoices</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mDefaultMedusaChoices</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">1</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">1</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">2</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">1</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">2</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">3</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">3</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">4</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">4</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">2</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">5</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">1</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">5</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">6</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">6</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">7</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">1</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">1</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">1</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">7</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">8</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">2</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">3</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">9</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">8</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">9</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">1</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">2</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">1</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">2</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">3</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">4</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">2</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">1</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">4</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">5</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">1</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">1</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">6</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">3</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">5</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">1</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">3</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">7</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">8</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">9</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">6</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">4</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">1</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">4</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">7</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">1</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">2</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">2</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">3</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">1</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">2</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">2</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">8</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">5</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">1</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">5</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">1</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">1</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">2</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">1</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">9</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">6</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">1</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">1</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">6</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">7</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">}</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12MedusaModule21mDefaultMedusaChoicesE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="memorycounters-h">
|
||
<h2>memoryCounters.h<a class="headerlink" href="#memorycounters-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14MemoryCountersE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14MemoryCountersE"></span><span id="_CPPv2N12tensorrt_llm7runtime14MemoryCountersE"></span><span id="tensorrt_llm::runtime::MemoryCounters"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">MemoryCounters</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCountersE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14MemoryCounters10SizeType32E">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14MemoryCounters10SizeType32E"></span><span id="_CPPv2N12tensorrt_llm7runtime14MemoryCounters10SizeType32E"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a7f97eec5349aa1601caef17d277d5f46"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">SizeType32</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters10SizeType32E" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14MemoryCounters8DiffTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14MemoryCounters8DiffTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14MemoryCounters8DiffTypeE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1ab5afaef89516ec27d7934229de37e387"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">DiffType</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ptrdiff_t</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters8DiffTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14MemoryCounters14MemoryCountersEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14MemoryCounters14MemoryCountersEv"></span><span id="_CPPv2N12tensorrt_llm7runtime14MemoryCounters14MemoryCountersEv"></span><span id="tensorrt_llm::runtime::MemoryCounters::MemoryCounters"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a97911bd1f11691ed4a652bf78e049e08"></span><span class="sig-name descname"><span class="n"><span class="pre">MemoryCounters</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">default</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters14MemoryCountersEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14MemoryCounters6getGpuEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14MemoryCounters6getGpuEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14MemoryCounters6getGpuEv"></span><span id="tensorrt_llm::runtime::MemoryCounters::getGpuC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1af804b97c2f575556afca28c10e0fd246"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters10SizeType32E" title="tensorrt_llm::runtime::MemoryCounters::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getGpu</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime14MemoryCounters6getGpuEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14MemoryCounters6getCpuEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14MemoryCounters6getCpuEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14MemoryCounters6getCpuEv"></span><span id="tensorrt_llm::runtime::MemoryCounters::getCpuC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1ac559f34eb120bea0f1c499997b6bb7eb"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters10SizeType32E" title="tensorrt_llm::runtime::MemoryCounters::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getCpu</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime14MemoryCounters6getCpuEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14MemoryCounters9getPinnedEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14MemoryCounters9getPinnedEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14MemoryCounters9getPinnedEv"></span><span id="tensorrt_llm::runtime::MemoryCounters::getPinnedC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a4e60cc67231f09228130d20cefb4de0f"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters10SizeType32E" title="tensorrt_llm::runtime::MemoryCounters::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getPinned</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime14MemoryCounters9getPinnedEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14MemoryCounters6getUVMEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14MemoryCounters6getUVMEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14MemoryCounters6getUVMEv"></span><span id="tensorrt_llm::runtime::MemoryCounters::getUVMC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a1fc4d4828c2838435fbf131d698d035f"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters10SizeType32E" title="tensorrt_llm::runtime::MemoryCounters::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getUVM</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime14MemoryCounters6getUVMEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14MemoryCounters13getPinnedPoolEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14MemoryCounters13getPinnedPoolEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14MemoryCounters13getPinnedPoolEv"></span><span id="tensorrt_llm::runtime::MemoryCounters::getPinnedPoolC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a0937904873a33859769278a94afcb79f"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters10SizeType32E" title="tensorrt_llm::runtime::MemoryCounters::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getPinnedPool</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime14MemoryCounters13getPinnedPoolEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14MemoryCounters10getGpuDiffEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14MemoryCounters10getGpuDiffEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14MemoryCounters10getGpuDiffEv"></span><span id="tensorrt_llm::runtime::MemoryCounters::getGpuDiffC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a399a7926f7819096b4f29fa04b3290de"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters8DiffTypeE" title="tensorrt_llm::runtime::MemoryCounters::DiffType"><span class="n"><span class="pre">DiffType</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getGpuDiff</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime14MemoryCounters10getGpuDiffEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14MemoryCounters10getCpuDiffEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14MemoryCounters10getCpuDiffEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14MemoryCounters10getCpuDiffEv"></span><span id="tensorrt_llm::runtime::MemoryCounters::getCpuDiffC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a62f4e6d3f0dd20c4ec6b35b4b6df0d47"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters8DiffTypeE" title="tensorrt_llm::runtime::MemoryCounters::DiffType"><span class="n"><span class="pre">DiffType</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getCpuDiff</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime14MemoryCounters10getCpuDiffEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14MemoryCounters13getPinnedDiffEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14MemoryCounters13getPinnedDiffEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14MemoryCounters13getPinnedDiffEv"></span><span id="tensorrt_llm::runtime::MemoryCounters::getPinnedDiffC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a4571ac85d01e0844b946d57615523ad0"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters8DiffTypeE" title="tensorrt_llm::runtime::MemoryCounters::DiffType"><span class="n"><span class="pre">DiffType</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getPinnedDiff</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime14MemoryCounters13getPinnedDiffEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14MemoryCounters10getUVMDiffEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14MemoryCounters10getUVMDiffEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14MemoryCounters10getUVMDiffEv"></span><span id="tensorrt_llm::runtime::MemoryCounters::getUVMDiffC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a29049162d689c1fc63d6df275d13c5d8"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters8DiffTypeE" title="tensorrt_llm::runtime::MemoryCounters::DiffType"><span class="n"><span class="pre">DiffType</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getUVMDiff</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime14MemoryCounters10getUVMDiffEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14MemoryCounters17getPinnedPoolDiffEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14MemoryCounters17getPinnedPoolDiffEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14MemoryCounters17getPinnedPoolDiffEv"></span><span id="tensorrt_llm::runtime::MemoryCounters::getPinnedPoolDiffC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a5a0c1d6212795781ae09a60cfadaa4b6"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters8DiffTypeE" title="tensorrt_llm::runtime::MemoryCounters::DiffType"><span class="n"><span class="pre">DiffType</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getPinnedPoolDiff</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime14MemoryCounters17getPinnedPoolDiffEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I_10MemoryTypeEN12tensorrt_llm7runtime14MemoryCounters8allocateEv10SizeType32">
|
||
<span id="_CPPv3I_10MemoryTypeEN12tensorrt_llm7runtime14MemoryCounters8allocateE10SizeType32"></span><span id="_CPPv2I_10MemoryTypeEN12tensorrt_llm7runtime14MemoryCounters8allocateE10SizeType32"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="tensorrt_llm::runtime::MemoryType"><span class="n"><span class="pre">MemoryType</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1abbd4ec1ac2e70b15f25f4a11fe28f99c"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">allocate</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters10SizeType32E" title="tensorrt_llm::runtime::MemoryCounters::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I_10MemoryTypeEN12tensorrt_llm7runtime14MemoryCounters8allocateEv10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14MemoryCounters8allocateE10MemoryType10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14MemoryCounters8allocateE10MemoryType10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime14MemoryCounters8allocateE10MemoryType10SizeType32"></span><span id="tensorrt_llm::runtime::MemoryCounters::allocate__MemoryType.SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a2d7815f9c1dd182a75924b8c3a7262db"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">allocate</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="tensorrt_llm::runtime::MemoryType"><span class="n"><span class="pre">MemoryType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">memoryType</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters10SizeType32E" title="tensorrt_llm::runtime::MemoryCounters::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters8allocateE10MemoryType10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I_10MemoryTypeEN12tensorrt_llm7runtime14MemoryCounters10deallocateEv10SizeType32">
|
||
<span id="_CPPv3I_10MemoryTypeEN12tensorrt_llm7runtime14MemoryCounters10deallocateE10SizeType32"></span><span id="_CPPv2I_10MemoryTypeEN12tensorrt_llm7runtime14MemoryCounters10deallocateE10SizeType32"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="tensorrt_llm::runtime::MemoryType"><span class="n"><span class="pre">MemoryType</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1aea6614ea6ce06f5c28d45bdf296de7f6"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">deallocate</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters10SizeType32E" title="tensorrt_llm::runtime::MemoryCounters::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I_10MemoryTypeEN12tensorrt_llm7runtime14MemoryCounters10deallocateEv10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14MemoryCounters10deallocateE10MemoryType10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14MemoryCounters10deallocateE10MemoryType10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime14MemoryCounters10deallocateE10MemoryType10SizeType32"></span><span id="tensorrt_llm::runtime::MemoryCounters::deallocate__MemoryType.SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a874d15b6d6c8f74f0a6ba37ff66d81c4"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">deallocate</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="tensorrt_llm::runtime::MemoryType"><span class="n"><span class="pre">MemoryType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">memoryType</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters10SizeType32E" title="tensorrt_llm::runtime::MemoryCounters::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters10deallocateE10MemoryType10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14MemoryCounters8toStringEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14MemoryCounters8toStringEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14MemoryCounters8toStringEv"></span><span id="tensorrt_llm::runtime::MemoryCounters::toStringC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1ad6981e0fc09bbeefdf4b2854fd03d77a"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">toString</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime14MemoryCounters8toStringEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-functions">Public Static Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14MemoryCounters11getInstanceEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14MemoryCounters11getInstanceEv"></span><span id="_CPPv2N12tensorrt_llm7runtime14MemoryCounters11getInstanceEv"></span><span id="tensorrt_llm::runtime::MemoryCounters::getInstance"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a71d4df0ab707ba942e7a0e5d18f092a5"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCountersE" title="tensorrt_llm::runtime::MemoryCounters"><span class="n"><span class="pre">MemoryCounters</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="n"><span class="pre">getInstance</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters11getInstanceEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14MemoryCounters13bytesToStringE10SizeType32i">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14MemoryCounters13bytesToStringE10SizeType32i"></span><span id="_CPPv2N12tensorrt_llm7runtime14MemoryCounters13bytesToStringE10SizeType32i"></span><span id="tensorrt_llm::runtime::MemoryCounters::bytesToString__SizeType32.i"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a31b1552f38254bdd173a8050321a57a5"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">bytesToString</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters10SizeType32E" title="tensorrt_llm::runtime::MemoryCounters::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">bytes</span></span>, <span class="kt"><span class="pre">int</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">precision</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">2</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters13bytesToStringE10SizeType32i" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14MemoryCounters13bytesToStringE8DiffTypei">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14MemoryCounters13bytesToStringE8DiffTypei"></span><span id="_CPPv2N12tensorrt_llm7runtime14MemoryCounters13bytesToStringE8DiffTypei"></span><span id="tensorrt_llm::runtime::MemoryCounters::bytesToString__DiffType.i"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1ae644945f3b01297e95a0f1724b5b4d8a"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">bytesToString</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters8DiffTypeE" title="tensorrt_llm::runtime::MemoryCounters::DiffType"><span class="n"><span class="pre">DiffType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">bytes</span></span>, <span class="kt"><span class="pre">int</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">precision</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">2</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters13bytesToStringE8DiffTypei" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-members">Private Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14MemoryCounters4mGpuE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14MemoryCounters4mGpuE"></span><span id="_CPPv2N12tensorrt_llm7runtime14MemoryCounters4mGpuE"></span><span id="tensorrt_llm::runtime::MemoryCounters::mGpu__std::atomic:SizeType32:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a3e0ea65103a976306619892dc6ebf16a"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">atomic</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters10SizeType32E" title="tensorrt_llm::runtime::MemoryCounters::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mGpu</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters4mGpuE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14MemoryCounters4mCpuE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14MemoryCounters4mCpuE"></span><span id="_CPPv2N12tensorrt_llm7runtime14MemoryCounters4mCpuE"></span><span id="tensorrt_llm::runtime::MemoryCounters::mCpu__std::atomic:SizeType32:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1aab2d4eec90ab57deb615843586f552f7"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">atomic</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters10SizeType32E" title="tensorrt_llm::runtime::MemoryCounters::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mCpu</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters4mCpuE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14MemoryCounters7mPinnedE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14MemoryCounters7mPinnedE"></span><span id="_CPPv2N12tensorrt_llm7runtime14MemoryCounters7mPinnedE"></span><span id="tensorrt_llm::runtime::MemoryCounters::mPinned__std::atomic:SizeType32:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1ad84cb0d4494a355210b3afad7fb5af7a"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">atomic</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters10SizeType32E" title="tensorrt_llm::runtime::MemoryCounters::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mPinned</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters7mPinnedE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14MemoryCounters4mUVME">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14MemoryCounters4mUVME"></span><span id="_CPPv2N12tensorrt_llm7runtime14MemoryCounters4mUVME"></span><span id="tensorrt_llm::runtime::MemoryCounters::mUVM__std::atomic:SizeType32:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a84274b5e6c96b46ad19ef29d80eece02"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">atomic</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters10SizeType32E" title="tensorrt_llm::runtime::MemoryCounters::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mUVM</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters4mUVME" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14MemoryCounters11mPinnedPoolE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14MemoryCounters11mPinnedPoolE"></span><span id="_CPPv2N12tensorrt_llm7runtime14MemoryCounters11mPinnedPoolE"></span><span id="tensorrt_llm::runtime::MemoryCounters::mPinnedPool__std::atomic:SizeType32:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a170062a3dc6bc26b4c95f526665d93e3"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">atomic</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters10SizeType32E" title="tensorrt_llm::runtime::MemoryCounters::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mPinnedPool</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters11mPinnedPoolE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14MemoryCounters8mGpuDiffE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14MemoryCounters8mGpuDiffE"></span><span id="_CPPv2N12tensorrt_llm7runtime14MemoryCounters8mGpuDiffE"></span><span id="tensorrt_llm::runtime::MemoryCounters::mGpuDiff__std::atomic:DiffType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a9f1e18f49dcefe7ad92c21cd7ec3a078"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">atomic</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters8DiffTypeE" title="tensorrt_llm::runtime::MemoryCounters::DiffType"><span class="n"><span class="pre">DiffType</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mGpuDiff</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters8mGpuDiffE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14MemoryCounters8mCpuDiffE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14MemoryCounters8mCpuDiffE"></span><span id="_CPPv2N12tensorrt_llm7runtime14MemoryCounters8mCpuDiffE"></span><span id="tensorrt_llm::runtime::MemoryCounters::mCpuDiff__std::atomic:DiffType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a442537225b82fc182cced193b9cc53bc"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">atomic</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters8DiffTypeE" title="tensorrt_llm::runtime::MemoryCounters::DiffType"><span class="n"><span class="pre">DiffType</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mCpuDiff</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters8mCpuDiffE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14MemoryCounters11mPinnedDiffE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14MemoryCounters11mPinnedDiffE"></span><span id="_CPPv2N12tensorrt_llm7runtime14MemoryCounters11mPinnedDiffE"></span><span id="tensorrt_llm::runtime::MemoryCounters::mPinnedDiff__std::atomic:DiffType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a8cfbf01ccdbe867c250798af6b41af81"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">atomic</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters8DiffTypeE" title="tensorrt_llm::runtime::MemoryCounters::DiffType"><span class="n"><span class="pre">DiffType</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mPinnedDiff</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters11mPinnedDiffE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14MemoryCounters8mUVMDiffE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14MemoryCounters8mUVMDiffE"></span><span id="_CPPv2N12tensorrt_llm7runtime14MemoryCounters8mUVMDiffE"></span><span id="tensorrt_llm::runtime::MemoryCounters::mUVMDiff__std::atomic:DiffType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1ab95b129271bda42cbb4f24545404896a"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">atomic</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters8DiffTypeE" title="tensorrt_llm::runtime::MemoryCounters::DiffType"><span class="n"><span class="pre">DiffType</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mUVMDiff</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters8mUVMDiffE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14MemoryCounters15mPinnedPoolDiffE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14MemoryCounters15mPinnedPoolDiffE"></span><span id="_CPPv2N12tensorrt_llm7runtime14MemoryCounters15mPinnedPoolDiffE"></span><span id="tensorrt_llm::runtime::MemoryCounters::mPinnedPoolDiff__std::atomic:DiffType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a0c7be43e6e4862f11811eae2979634c4"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">atomic</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters8DiffTypeE" title="tensorrt_llm::runtime::MemoryCounters::DiffType"><span class="n"><span class="pre">DiffType</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mPinnedPoolDiff</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters15mPinnedPoolDiffE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="prompttuningparams-h">
|
||
<h2>promptTuningParams.h<a class="headerlink" href="#prompttuningparams-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime25GenericPromptTuningParamsE">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime25GenericPromptTuningParamsE"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime25GenericPromptTuningParamsE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TTensor</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericPromptTuningParams"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">GenericPromptTuningParams</span></span></span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime25GenericPromptTuningParamsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime25GenericPromptTuningParams9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime25GenericPromptTuningParams9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericPromptTuningParams_1a7f0226e537f0c4a164c298adf9731e67"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime25GenericPromptTuningParamsE" title="tensorrt_llm::runtime::GenericPromptTuningParams::TTensor"><span class="n"><span class="pre">TTensor</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams9TensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams10SizeType32E">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime25GenericPromptTuningParams10SizeType32E"></span><span id="_CPPv2N12tensorrt_llm7runtime25GenericPromptTuningParams10SizeType32E"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericPromptTuningParams_1a91aed2fabb50029ec34af4a22a952b77"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">SizeType32</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams10SizeType32E" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams25GenericPromptTuningParamsE9TensorPtr9TensorPtr9TensorPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime25GenericPromptTuningParams25GenericPromptTuningParamsE9TensorPtr9TensorPtr9TensorPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime25GenericPromptTuningParams25GenericPromptTuningParamsE9TensorPtr9TensorPtr9TensorPtr"></span><span id="tensorrt_llm::runtime::GenericPromptTuningParams::GenericPromptTuningParams__TensorPtr.TensorPtr.TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericPromptTuningParams_1a08b54a1a315a75cada66405e31709668"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">GenericPromptTuningParams</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams9TensorPtrE" title="tensorrt_llm::runtime::GenericPromptTuningParams::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">embeddingTable</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams9TensorPtrE" title="tensorrt_llm::runtime::GenericPromptTuningParams::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="p"><span class="pre">(</span></span><span class="p"><span class="pre">)</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams9TensorPtrE" title="tensorrt_llm::runtime::GenericPromptTuningParams::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tasks</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams9TensorPtrE" title="tensorrt_llm::runtime::GenericPromptTuningParams::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="p"><span class="pre">(</span></span><span class="p"><span class="pre">)</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams9TensorPtrE" title="tensorrt_llm::runtime::GenericPromptTuningParams::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">vocabSize</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams9TensorPtrE" title="tensorrt_llm::runtime::GenericPromptTuningParams::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="p"><span class="pre">(</span></span><span class="p"><span class="pre">)</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams25GenericPromptTuningParamsE9TensorPtr9TensorPtr9TensorPtr" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams14embeddingTableE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime25GenericPromptTuningParams14embeddingTableE"></span><span id="_CPPv2N12tensorrt_llm7runtime25GenericPromptTuningParams14embeddingTableE"></span><span id="tensorrt_llm::runtime::GenericPromptTuningParams::embeddingTable__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericPromptTuningParams_1a718fcc98779017166cbe0a748c081414"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams9TensorPtrE" title="tensorrt_llm::runtime::GenericPromptTuningParams::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">embeddingTable</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams14embeddingTableE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams5tasksE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime25GenericPromptTuningParams5tasksE"></span><span id="_CPPv2N12tensorrt_llm7runtime25GenericPromptTuningParams5tasksE"></span><span id="tensorrt_llm::runtime::GenericPromptTuningParams::tasks__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericPromptTuningParams_1aea7aafec3a272782a1a467eb159af6e1"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams9TensorPtrE" title="tensorrt_llm::runtime::GenericPromptTuningParams::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tasks</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams5tasksE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams9vocabSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime25GenericPromptTuningParams9vocabSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime25GenericPromptTuningParams9vocabSizeE"></span><span id="tensorrt_llm::runtime::GenericPromptTuningParams::vocabSize__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericPromptTuningParams_1acb732cd787508ea7013cf1e9002590ea"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams9TensorPtrE" title="tensorrt_llm::runtime::GenericPromptTuningParams::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">vocabSize</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams9vocabSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams19promptTuningEnabledE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime25GenericPromptTuningParams19promptTuningEnabledE"></span><span id="_CPPv2N12tensorrt_llm7runtime25GenericPromptTuningParams19promptTuningEnabledE"></span><span id="tensorrt_llm::runtime::GenericPromptTuningParams::promptTuningEnabled__std::vector:b:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenericPromptTuningParams_1a4f994aa4122125396dee70846aec1318"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="kt"><span class="pre">bool</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">promptTuningEnabled</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams19promptTuningEnabledE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime18PromptTuningParamsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime18PromptTuningParamsE"></span><span id="_CPPv2N12tensorrt_llm7runtime18PromptTuningParamsE"></span><span id="tensorrt_llm::runtime::PromptTuningParams"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1PromptTuningParams"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">PromptTuningParams</span></span></span><span class="w"> </span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="k"><span class="pre">public</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime25GenericPromptTuningParamsE" title="tensorrt_llm::runtime::GenericPromptTuningParams"><span class="n"><span class="pre">GenericPromptTuningParams</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime18PromptTuningParamsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime18PromptTuningParams9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime18PromptTuningParams9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime18PromptTuningParams9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1PromptTuningParams_1acf57960de488b9fcc3021a34690f003a"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime18PromptTuningParams9TensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime18PromptTuningParams10SizeType32E">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime18PromptTuningParams10SizeType32E"></span><span id="_CPPv2N12tensorrt_llm7runtime18PromptTuningParams10SizeType32E"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1PromptTuningParams_1a99d9372e247512429412043f3103e08a"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">SizeType32</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime25GenericPromptTuningParamsE" title="tensorrt_llm::runtime::GenericPromptTuningParams"><span class="n"><span class="pre">GenericPromptTuningParams</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25GenericPromptTuningParams10SizeType32E" title="tensorrt_llm::runtime::GenericPromptTuningParams::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime18PromptTuningParams10SizeType32E" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime18PromptTuningParams18PromptTuningParamsE9TensorPtr9TensorPtr9TensorPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime18PromptTuningParams18PromptTuningParamsE9TensorPtr9TensorPtr9TensorPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime18PromptTuningParams18PromptTuningParamsE9TensorPtr9TensorPtr9TensorPtr"></span><span id="tensorrt_llm::runtime::PromptTuningParams::PromptTuningParams__TensorPtr.TensorPtr.TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1PromptTuningParams_1a6439bfa322d7dc1f589cd3877640a952"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">PromptTuningParams</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18PromptTuningParams9TensorPtrE" title="tensorrt_llm::runtime::PromptTuningParams::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">embeddingTable</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">nullptr</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18PromptTuningParams9TensorPtrE" title="tensorrt_llm::runtime::PromptTuningParams::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tasks</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">nullptr</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18PromptTuningParams9TensorPtrE" title="tensorrt_llm::runtime::PromptTuningParams::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">vocabSize</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">nullptr</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime18PromptTuningParams18PromptTuningParamsE9TensorPtr9TensorPtr9TensorPtr" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime18PromptTuningParams15fillTasksTensorE9TensorPtrK10SizeType32K10SizeType32RKNSt6vectorI10SizeType32EERKNSt6vectorI10SizeType32EERK13BufferManagerb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime18PromptTuningParams15fillTasksTensorE9TensorPtrK10SizeType32K10SizeType32RKNSt6vectorI10SizeType32EERKNSt6vectorI10SizeType32EERK13BufferManagerb"></span><span id="_CPPv2N12tensorrt_llm7runtime18PromptTuningParams15fillTasksTensorE9TensorPtrK10SizeType32K10SizeType32RKNSt6vectorI10SizeType32EERKNSt6vectorI10SizeType32EERK13BufferManagerb"></span><span id="tensorrt_llm::runtime::PromptTuningParams::fillTasksTensor__TensorPtr.SizeType32C.SizeType32C.std::vector:SizeType32:CR.std::vector:SizeType32:CR.BufferManagerCR.b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1PromptTuningParams_1afb433c94f2582aa98fbb2080adafd0f9"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">fillTasksTensor</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18PromptTuningParams9TensorPtrE" title="tensorrt_llm::runtime::PromptTuningParams::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tasksHost</span></span>, <span class="k"><span class="pre">const</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18PromptTuningParams10SizeType32E" title="tensorrt_llm::runtime::PromptTuningParams::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchSize</span></span>, <span class="k"><span class="pre">const</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18PromptTuningParams10SizeType32E" title="tensorrt_llm::runtime::PromptTuningParams::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">numContextRequests</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18PromptTuningParams10SizeType32E" title="tensorrt_llm::runtime::PromptTuningParams::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">reqBeamWidths</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18PromptTuningParams10SizeType32E" title="tensorrt_llm::runtime::PromptTuningParams::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">reqPromptLengths</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManagerE" title="tensorrt_llm::runtime::BufferManager"><span class="n"><span class="pre">BufferManager</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">manager</span></span>, <span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">packedInput</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime18PromptTuningParams15fillTasksTensorE9TensorPtrK10SizeType32K10SizeType32RKNSt6vectorI10SizeType32EERKNSt6vectorI10SizeType32EERK13BufferManagerb" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="rawengine-h">
|
||
<h2>rawEngine.h<a class="headerlink" href="#rawengine-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9RawEngineE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9RawEngineE"></span><span id="_CPPv2N12tensorrt_llm7runtime9RawEngineE"></span><span id="tensorrt_llm::runtime::RawEngine"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1RawEngine"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">RawEngine</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9RawEngineE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp enum">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9RawEngine4TypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9RawEngine4TypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime9RawEngine4TypeE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1RawEngine_1a3c8f3acf0c858144e0987346247af075"></span><span class="k"><span class="pre">enum</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Type</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine4TypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p><em>Values:</em></p>
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9RawEngine4Type8FilePathE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9RawEngine4Type8FilePathE"></span><span id="_CPPv2N12tensorrt_llm7runtime9RawEngine4Type8FilePathE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1RawEngine_1a3c8f3acf0c858144e0987346247af075a05f8d2398fcf614f8784248055f32e17"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">FilePath</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine4Type8FilePathE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9RawEngine4Type15AddressWithSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9RawEngine4Type15AddressWithSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime9RawEngine4Type15AddressWithSizeE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1RawEngine_1a3c8f3acf0c858144e0987346247af075a05dee9767a15cf70383d2faf6974afe9"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">AddressWithSize</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine4Type15AddressWithSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9RawEngine4Type10HostMemoryE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9RawEngine4Type10HostMemoryE"></span><span id="_CPPv2N12tensorrt_llm7runtime9RawEngine4Type10HostMemoryE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1RawEngine_1a3c8f3acf0c858144e0987346247af075aff64ba05059375611eb50ec057d3996f"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">HostMemory</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine4Type10HostMemoryE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9RawEngine9RawEngineENSt10filesystem4pathE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9RawEngine9RawEngineENSt10filesystem4pathE"></span><span id="_CPPv2N12tensorrt_llm7runtime9RawEngine9RawEngineENSt10filesystem4pathE"></span><span id="tensorrt_llm::runtime::RawEngine::RawEngine__std::filesystem::path"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1RawEngine_1adea0c457ec4dcaef069c6d9cb218014b"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">RawEngine</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">filesystem</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">path</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">enginePath</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine9RawEngineENSt10filesystem4pathE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9RawEngine9RawEngineEPKvNSt6size_tE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9RawEngine9RawEngineEPKvNSt6size_tE"></span><span id="_CPPv2N12tensorrt_llm7runtime9RawEngine9RawEngineEPKvNSt6size_tE"></span><span id="tensorrt_llm::runtime::RawEngine::RawEngine__voidCP.std::s"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1RawEngine_1a3052697810402d1db99e6c5a4da1bb5a"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">RawEngine</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="n sig-param"><span class="pre">engineAddr</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">engineSize</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine9RawEngineEPKvNSt6size_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9RawEngine9RawEngineEPKN8nvinfer111IHostMemoryE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9RawEngine9RawEngineEPKN8nvinfer111IHostMemoryE"></span><span id="_CPPv2N12tensorrt_llm7runtime9RawEngine9RawEngineEPKN8nvinfer111IHostMemoryE"></span><span id="tensorrt_llm::runtime::RawEngine::RawEngine__nvinfer1::IHostMemoryCP"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1RawEngine_1abdaa133b12d2a8a8ff00ee6ac81b6d67"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">RawEngine</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">IHostMemory</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="n sig-param"><span class="pre">engineBuffer</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine9RawEngineEPKN8nvinfer111IHostMemoryE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime9RawEngine7getTypeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime9RawEngine7getTypeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime9RawEngine7getTypeEv"></span><span id="tensorrt_llm::runtime::RawEngine::getTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1RawEngine_1a77dc57a1fbc1a105a2cf206e030ad1bf"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine4TypeE" title="tensorrt_llm::runtime::RawEngine::Type"><span class="n"><span class="pre">Type</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getType</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime9RawEngine7getTypeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime9RawEngine7getPathEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime9RawEngine7getPathEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime9RawEngine7getPathEv"></span><span id="tensorrt_llm::runtime::RawEngine::getPathC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1RawEngine_1a737a7c9d9876ad0cc3b21cf5926a5787"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">filesystem</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">path</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getPath</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime9RawEngine7getPathEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime9RawEngine10getPathOptEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime9RawEngine10getPathOptEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime9RawEngine10getPathOptEv"></span><span id="tensorrt_llm::runtime::RawEngine::getPathOptC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1RawEngine_1a420b0935cefb09c44966918cec37523b"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">filesystem</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">path</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getPathOpt</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime9RawEngine10getPathOptEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9RawEngine7setPathENSt10filesystem4pathE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9RawEngine7setPathENSt10filesystem4pathE"></span><span id="_CPPv2N12tensorrt_llm7runtime9RawEngine7setPathENSt10filesystem4pathE"></span><span id="tensorrt_llm::runtime::RawEngine::setPath__std::filesystem::path"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1RawEngine_1a29fa3156e7ef0efc344cef39be7536f9"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setPath</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">filesystem</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">path</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">enginePath</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine7setPathENSt10filesystem4pathE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime9RawEngine23getManagedWeightsMapOptEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime9RawEngine23getManagedWeightsMapOptEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime9RawEngine23getManagedWeightsMapOptEv"></span><span id="tensorrt_llm::runtime::RawEngine::getManagedWeightsMapOptC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1RawEngine_1a49ff9f55b13ee8dcb8b240b39f9fe4ab"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">map</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executorE" title="tensorrt_llm::executor"><span class="n"><span class="pre">executor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executor6TensorE" title="tensorrt_llm::executor::Tensor"><span class="n"><span class="pre">Tensor</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="n"><span class="pre">getManagedWeightsMapOpt</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime9RawEngine23getManagedWeightsMapOptEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9RawEngine20setManagedWeightsMapENSt3mapINSt6stringEN12tensorrt_llm8executor6TensorEEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9RawEngine20setManagedWeightsMapENSt3mapINSt6stringEN12tensorrt_llm8executor6TensorEEE"></span><span id="_CPPv2N12tensorrt_llm7runtime9RawEngine20setManagedWeightsMapENSt3mapINSt6stringEN12tensorrt_llm8executor6TensorEEE"></span><span id="tensorrt_llm::runtime::RawEngine::setManagedWeightsMap__std::map:ss.tensorrt_llm::executor::Tensor:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1RawEngine_1ab38ef50e4e4107e0fa3344a33787f210"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setManagedWeightsMap</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">map</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executorE" title="tensorrt_llm::executor"><span class="n"><span class="pre">executor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executor6TensorE" title="tensorrt_llm::executor::Tensor"><span class="n"><span class="pre">Tensor</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="n sig-param"><span class="pre">managedWeightsMap</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine20setManagedWeightsMapENSt3mapINSt6stringEN12tensorrt_llm8executor6TensorEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime9RawEngine10getAddressEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime9RawEngine10getAddressEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime9RawEngine10getAddressEv"></span><span id="tensorrt_llm::runtime::RawEngine::getAddressC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1RawEngine_1a9802245035ff5d5b1e15fddf08d47b7a"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="sig-name descname"><span class="n"><span class="pre">getAddress</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime9RawEngine10getAddressEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime9RawEngine7getSizeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime9RawEngine7getSizeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime9RawEngine7getSizeEv"></span><span id="tensorrt_llm::runtime::RawEngine::getSizeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1RawEngine_1a111cdd7dd515f0692199ae815f0aa186"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getSize</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime9RawEngine7getSizeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime9RawEngine13getHostMemoryEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime9RawEngine13getHostMemoryEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime9RawEngine13getHostMemoryEv"></span><span id="tensorrt_llm::runtime::RawEngine::getHostMemoryC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1RawEngine_1a82d55c1942b25bbad111fff71336066e"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">IHostMemory</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="sig-name descname"><span class="n"><span class="pre">getHostMemory</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime9RawEngine13getHostMemoryEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9RawEngine11mEngineAddrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9RawEngine11mEngineAddrE"></span><span id="_CPPv2N12tensorrt_llm7runtime9RawEngine11mEngineAddrE"></span><span id="tensorrt_llm::runtime::RawEngine::mEngineAddr__voidCP"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1RawEngine_1a95a24a9a7ada011d7f6260f5948098a1"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="sig-name descname"><span class="n"><span class="pre">mEngineAddr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine11mEngineAddrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9RawEngine11mEngineSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9RawEngine11mEngineSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime9RawEngine11mEngineSizeE"></span><span id="tensorrt_llm::runtime::RawEngine::mEngineSize__std::s"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1RawEngine_1a3c29c2fc9109c3155a75346a3ba4b859"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mEngineSize</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine11mEngineSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-members">Private Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9RawEngine5mTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9RawEngine5mTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime9RawEngine5mTypeE"></span><span id="tensorrt_llm::runtime::RawEngine::mType__Type"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1RawEngine_1ae7f9d965ecc951e7eab03a556d5079ac"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine4TypeE" title="tensorrt_llm::runtime::RawEngine::Type"><span class="n"><span class="pre">Type</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mType</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine5mTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9RawEngine11mEnginePathE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9RawEngine11mEnginePathE"></span><span id="_CPPv2N12tensorrt_llm7runtime9RawEngine11mEnginePathE"></span><span id="tensorrt_llm::runtime::RawEngine::mEnginePath__std::optional:std::filesystem::path:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1RawEngine_1ae0a8b43561f46764487b26a60956a44d"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">filesystem</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">path</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mEnginePath</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine11mEnginePathE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="classtensorrt__llm_1_1runtime_1_1RawEngine_1a4b1eb35ae9e6aad50395b67a1d91c04f"></span><span class="sig-name descname"><span class="pre">struct</span> <span class="pre">tensorrt_llm::runtime::RawEngine</span></span></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9RawEngine13mEngineBufferE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9RawEngine13mEngineBufferE"></span><span id="_CPPv2N12tensorrt_llm7runtime9RawEngine13mEngineBufferE"></span><span id="tensorrt_llm::runtime::RawEngine::mEngineBuffer__nvinfer1::IHostMemoryCP"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1RawEngine_1a991cf95f70001c61a7062bf62a1684ae"></span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">IHostMemory</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="sig-name descname"><span class="n"><span class="pre">mEngineBuffer</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine13mEngineBufferE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9RawEngine18mManagedWeightsMapE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9RawEngine18mManagedWeightsMapE"></span><span id="_CPPv2N12tensorrt_llm7runtime9RawEngine18mManagedWeightsMapE"></span><span id="tensorrt_llm::runtime::RawEngine::mManagedWeightsMap__std::optional:std::map:ss.tensorrt_llm::executor::Tensor::"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1RawEngine_1a48f7d27244eacf245b02d90ded15f83e"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">map</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executorE" title="tensorrt_llm::executor"><span class="n"><span class="pre">executor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executor6TensorE" title="tensorrt_llm::executor::Tensor"><span class="n"><span class="pre">Tensor</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mManagedWeightsMap</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9RawEngine18mManagedWeightsMapE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="request-h">
|
||
<h2>request.h<a class="headerlink" href="#request-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batchE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batchE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batchE"></span><span id="tensorrt_llm::runtime::decoder_batch"></span><span class="target" id="namespacetensorrt__llm_1_1runtime_1_1decoder__batch"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">decoder_batch</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch7RequestE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch7RequestE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch7RequestE"></span><span id="tensorrt_llm::runtime::decoder_batch::Request"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Request"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Request</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7RequestE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request14TensorConstPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch7Request14TensorConstPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch7Request14TensorConstPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Request_1a6d5f16a0733fae53d2f6a74805cb012b"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorConstPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor14SharedConstPtrE" title="tensorrt_llm::runtime::ITensor::SharedConstPtr"><span class="n"><span class="pre">SharedConstPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request14TensorConstPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch7Request9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch7Request9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Request_1a95d9dd610e7a6295d06f75a33cbb4c43"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request9TensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request9BufferPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch7Request9BufferPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch7Request9BufferPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Request_1a78becbf7fa5928420a994c0d9d9ddb65"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">BufferPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer9SharedPtrE" title="tensorrt_llm::runtime::IBuffer::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request9BufferPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request7RequestE14TensorConstPtr10SizeType32NSt8optionalI10SizeType32EENSt8optionalI10SizeType32EE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch7Request7RequestE14TensorConstPtr10SizeType32NSt8optionalI10SizeType32EENSt8optionalI10SizeType32EE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch7Request7RequestE14TensorConstPtr10SizeType32NSt8optionalI10SizeType32EENSt8optionalI10SizeType32EE"></span><span id="tensorrt_llm::runtime::decoder_batch::Request::Request__TensorConstPtr.SizeType32.std::optional:SizeType32:.std::optional:SizeType32:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Request_1a0dad449c01fb8b4af63de85659096fe0"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Request</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request14TensorConstPtrE" title="tensorrt_llm::runtime::decoder_batch::Request::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">ids</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">inputLen</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="n sig-param"><span class="pre">maxNewTokens</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">nullopt</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="n sig-param"><span class="pre">endId</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">nullopt</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request7RequestE14TensorConstPtr10SizeType32NSt8optionalI10SizeType32EENSt8optionalI10SizeType32EE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request3idsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch7Request3idsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch7Request3idsE"></span><span id="tensorrt_llm::runtime::decoder_batch::Request::ids__TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Request_1a3e55abb4aecb8f67610629a44c74ae08"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request14TensorConstPtrE" title="tensorrt_llm::runtime::decoder_batch::Request::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">ids</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request3idsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request8inputLenE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch7Request8inputLenE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch7Request8inputLenE"></span><span id="tensorrt_llm::runtime::decoder_batch::Request::inputLen__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Request_1a8762d6cd8c7a7efecb04d08c1d28e837"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">inputLen</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request8inputLenE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request12maxNewTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch7Request12maxNewTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch7Request12maxNewTokensE"></span><span id="tensorrt_llm::runtime::decoder_batch::Request::maxNewTokens__std::optional:SizeType32:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Request_1a5149bb633e0daa432cecc3d7c50483d3"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">maxNewTokens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request12maxNewTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request5endIdE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch7Request5endIdE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch7Request5endIdE"></span><span id="tensorrt_llm::runtime::decoder_batch::Request::endId__std::optional:SizeType32:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Request_1a9ab7466963ecf94768b4ecdc00702973"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">endId</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request5endIdE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request11draftTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch7Request11draftTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch7Request11draftTokensE"></span><span id="tensorrt_llm::runtime::decoder_batch::Request::draftTokens__BufferPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Request_1a1a524b5348e2d8ea442050f50ff3f1c7"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request9BufferPtrE" title="tensorrt_llm::runtime::decoder_batch::Request::BufferPtr"><span class="n"><span class="pre">BufferPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">draftTokens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request11draftTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request11draftLogitsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch7Request11draftLogitsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch7Request11draftLogitsE"></span><span id="tensorrt_llm::runtime::decoder_batch::Request::draftLogits__std::optional:TensorPtr:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Request_1afa9181bc52e25fe7dc0e925c69dc8d86"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request9TensorPtrE" title="tensorrt_llm::runtime::decoder_batch::Request::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">draftLogits</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request11draftLogitsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request13embeddingBiasE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch7Request13embeddingBiasE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch7Request13embeddingBiasE"></span><span id="tensorrt_llm::runtime::decoder_batch::Request::embeddingBias__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Request_1acbedf401f2f77534aa355f3ecbb0b755"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request9TensorPtrE" title="tensorrt_llm::runtime::decoder_batch::Request::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">embeddingBias</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request13embeddingBiasE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request12badWordsListE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch7Request12badWordsListE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch7Request12badWordsListE"></span><span id="tensorrt_llm::runtime::decoder_batch::Request::badWordsList__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Request_1af18cc1700b613308c14d492e69ab76e9"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request9TensorPtrE" title="tensorrt_llm::runtime::decoder_batch::Request::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">badWordsList</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request12badWordsListE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request13stopWordsListE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch7Request13stopWordsListE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch7Request13stopWordsListE"></span><span id="tensorrt_llm::runtime::decoder_batch::Request::stopWordsList__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Request_1addd0a66af57e7186287ac7e5853c0c9d"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request9TensorPtrE" title="tensorrt_llm::runtime::decoder_batch::Request::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">stopWordsList</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request13stopWordsListE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request28generatedTokensPerEngineStepE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch7Request28generatedTokensPerEngineStepE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch7Request28generatedTokensPerEngineStepE"></span><span id="tensorrt_llm::runtime::decoder_batch::Request::generatedTokensPerEngineStep__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Request_1ad0961016791825df1e39e2d4c8bced7a"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">generatedTokensPerEngineStep</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request28generatedTokensPerEngineStepE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request11medusaPathsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch7Request11medusaPathsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch7Request11medusaPathsE"></span><span id="tensorrt_llm::runtime::decoder_batch::Request::medusaPaths__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Request_1a2095b0e38974e5704d49f7b7226d1d62"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request9TensorPtrE" title="tensorrt_llm::runtime::decoder_batch::Request::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">medusaPaths</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request11medusaPathsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request13medusaTreeIdsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch7Request13medusaTreeIdsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch7Request13medusaTreeIdsE"></span><span id="tensorrt_llm::runtime::decoder_batch::Request::medusaTreeIds__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Request_1a53cd490ea4a4acc421b66a24ede31697"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request9TensorPtrE" title="tensorrt_llm::runtime::decoder_batch::Request::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">medusaTreeIds</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request13medusaTreeIdsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request22lookaheadRuntimeConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch7Request22lookaheadRuntimeConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch7Request22lookaheadRuntimeConfigE"></span><span id="tensorrt_llm::runtime::decoder_batch::Request::lookaheadRuntimeConfig__std::optional:executor::LookaheadDecodingConfig:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Request_1a724413e68cfc7bea981a1b1b334a1704"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executorE" title="tensorrt_llm::executor"><span class="n"><span class="pre">executor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executor23LookaheadDecodingConfigE" title="tensorrt_llm::executor::LookaheadDecodingConfig"><span class="n"><span class="pre">LookaheadDecodingConfig</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">lookaheadRuntimeConfig</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request22lookaheadRuntimeConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request11eagleConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch7Request11eagleConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch7Request11eagleConfigE"></span><span id="tensorrt_llm::runtime::decoder_batch::Request::eagleConfig__std::optional:executor::EagleConfig:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Request_1aaa297b6687699e8f43792997f503bef0"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executorE" title="tensorrt_llm::executor"><span class="n"><span class="pre">executor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executor11EagleConfigE" title="tensorrt_llm::executor::EagleConfig"><span class="n"><span class="pre">EagleConfig</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">eagleConfig</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request11eagleConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request5dtypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch7Request5dtypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch7Request5dtypeE"></span><span id="tensorrt_llm::runtime::decoder_batch::Request::dtype__nvinfer1::DataType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Request_1a3b1b508a767e8f4af619a2553239319b"></span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">dtype</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request5dtypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="runtimedefaults-h">
|
||
<h2>runtimeDefaults.h<a class="headerlink" href="#runtimedefaults-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15RuntimeDefaultsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15RuntimeDefaultsE"></span><span id="_CPPv2N12tensorrt_llm7runtime15RuntimeDefaultsE"></span><span id="tensorrt_llm::runtime::RuntimeDefaults"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1RuntimeDefaults"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">RuntimeDefaults</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15RuntimeDefaultsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15RuntimeDefaults15RuntimeDefaultsENSt8optionalINSt6vectorI10SizeType32EEEENSt8optionalI10SizeType32EE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15RuntimeDefaults15RuntimeDefaultsENSt8optionalINSt6vectorI10SizeType32EEEENSt8optionalI10SizeType32EE"></span><span id="_CPPv2N12tensorrt_llm7runtime15RuntimeDefaults15RuntimeDefaultsENSt8optionalINSt6vectorI10SizeType32EEEENSt8optionalI10SizeType32EE"></span><span id="tensorrt_llm::runtime::RuntimeDefaults::RuntimeDefaults__std::optional:std::vector:SizeType32::.std::optional:SizeType32:"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1RuntimeDefaults_1a4fbd807ecfe1abc3d6747ce3316885d3"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">RuntimeDefaults</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="n sig-param"><span class="pre">maxAttentionWindowVec</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="n sig-param"><span class="pre">sinkTokenLength</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15RuntimeDefaults15RuntimeDefaultsENSt8optionalINSt6vectorI10SizeType32EEEENSt8optionalI10SizeType32EE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15RuntimeDefaults15RuntimeDefaultsEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15RuntimeDefaults15RuntimeDefaultsEv"></span><span id="_CPPv2N12tensorrt_llm7runtime15RuntimeDefaults15RuntimeDefaultsEv"></span><span id="tensorrt_llm::runtime::RuntimeDefaults::RuntimeDefaults"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1RuntimeDefaults_1ad97400de900b8024bb80c4efc48aae88"></span><span class="sig-name descname"><span class="n"><span class="pre">RuntimeDefaults</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">default</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15RuntimeDefaults15RuntimeDefaultsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15RuntimeDefaults21maxAttentionWindowVecE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15RuntimeDefaults21maxAttentionWindowVecE"></span><span id="_CPPv2N12tensorrt_llm7runtime15RuntimeDefaults21maxAttentionWindowVecE"></span><span id="tensorrt_llm::runtime::RuntimeDefaults::maxAttentionWindowVec__std::optional:std::vector:SizeType32::"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1RuntimeDefaults_1ae1ccb7d93441677add1623e581440f40"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">maxAttentionWindowVec</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15RuntimeDefaults21maxAttentionWindowVecE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15RuntimeDefaults15sinkTokenLengthE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15RuntimeDefaults15sinkTokenLengthE"></span><span id="_CPPv2N12tensorrt_llm7runtime15RuntimeDefaults15sinkTokenLengthE"></span><span id="tensorrt_llm::runtime::RuntimeDefaults::sinkTokenLength__std::optional:SizeType32:"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1RuntimeDefaults_1aaea0e369a2d34c1b4e64d614281aeec3"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">sinkTokenLength</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15RuntimeDefaults15sinkTokenLengthE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="speculativedecodingmode-h">
|
||
<h2>speculativeDecodingMode.h<a class="headerlink" href="#speculativedecodingmode-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingModeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23SpeculativeDecodingModeE"></span><span id="_CPPv2N12tensorrt_llm7runtime23SpeculativeDecodingModeE"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">SpeculativeDecodingMode</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingModeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode14UnderlyingTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23SpeculativeDecodingMode14UnderlyingTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime23SpeculativeDecodingMode14UnderlyingTypeE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1a142fe60d488053b88f9961e51993cd4c"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">UnderlyingType</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">uint8_t</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode14UnderlyingTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode6isNoneEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime23SpeculativeDecodingMode6isNoneEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime23SpeculativeDecodingMode6isNoneEv"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::isNoneCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1acd3c978d723e3a4888d10f06f71adab6"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">isNone</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode6isNoneEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode21isDraftTokensExternalEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime23SpeculativeDecodingMode21isDraftTokensExternalEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime23SpeculativeDecodingMode21isDraftTokensExternalEv"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::isDraftTokensExternalCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1afee189c9b551928bb2645c14a8063871"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">isDraftTokensExternal</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode21isDraftTokensExternalEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode8isMedusaEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime23SpeculativeDecodingMode8isMedusaEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime23SpeculativeDecodingMode8isMedusaEv"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::isMedusaCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1ac386b84b5b4d90fd2bcc311514428c4d"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">isMedusa</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode8isMedusaEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode19isLookaheadDecodingEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime23SpeculativeDecodingMode19isLookaheadDecodingEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime23SpeculativeDecodingMode19isLookaheadDecodingEv"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::isLookaheadDecodingCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1a14e479f01ad1d809786603b6f1265b0f"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">isLookaheadDecoding</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode19isLookaheadDecodingEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode21isExplicitDraftTokensEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime23SpeculativeDecodingMode21isExplicitDraftTokensEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime23SpeculativeDecodingMode21isExplicitDraftTokensEv"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::isExplicitDraftTokensCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1aca96f1a5e256bc1e851819c44825ae02"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">isExplicitDraftTokens</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode21isExplicitDraftTokensEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode7isEagleEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime23SpeculativeDecodingMode7isEagleEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime23SpeculativeDecodingMode7isEagleEv"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::isEagleCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1a96e3d17ad9aa9a93df5e2e8cf029710a"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">isEagle</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode7isEagleEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode18updatesPositionIdsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime23SpeculativeDecodingMode18updatesPositionIdsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime23SpeculativeDecodingMode18updatesPositionIdsEv"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::updatesPositionIdsCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1a3ebdee2a9c4aebd54efa347ac4b48d33"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">updatesPositionIds</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode18updatesPositionIdsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode21requiresAttentionMaskEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime23SpeculativeDecodingMode21requiresAttentionMaskEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime23SpeculativeDecodingMode21requiresAttentionMaskEv"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::requiresAttentionMaskCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1adc3b3407372f26c7caf42f09ad5457a8"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">requiresAttentionMask</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode21requiresAttentionMaskEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode19predictsDraftTokensEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime23SpeculativeDecodingMode19predictsDraftTokensEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime23SpeculativeDecodingMode19predictsDraftTokensEv"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::predictsDraftTokensCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1a546d33fe11fea48eea7242d4d4279060"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">predictsDraftTokens</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode19predictsDraftTokensEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode18needsKVCacheRewindEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime23SpeculativeDecodingMode18needsKVCacheRewindEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime23SpeculativeDecodingMode18needsKVCacheRewindEv"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::needsKVCacheRewindCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1adc46b81262fd8e85146857dee395a438"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">needsKVCacheRewind</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode18needsKVCacheRewindEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode19variableDraftLengthEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime23SpeculativeDecodingMode19variableDraftLengthEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime23SpeculativeDecodingMode19variableDraftLengthEv"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::variableDraftLengthCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1a1fbbcb52a29d01aa4326fb6587502539"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">variableDraftLength</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode19variableDraftLengthEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode14hasDraftLogitsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime23SpeculativeDecodingMode14hasDraftLogitsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime23SpeculativeDecodingMode14hasDraftLogitsEv"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::hasDraftLogitsCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1a013f7fdcadf107956d33adcde8ad38f2"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">hasDraftLogits</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode14hasDraftLogitsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode20needsDecoderPrologueEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime23SpeculativeDecodingMode20needsDecoderPrologueEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime23SpeculativeDecodingMode20needsDecoderPrologueEv"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::needsDecoderPrologueCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1a9153eb4a918de5a7e093e426888d3986"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">needsDecoderPrologue</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode20needsDecoderPrologueEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingModeeqERK23SpeculativeDecodingMode">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime23SpeculativeDecodingModeeqERK23SpeculativeDecodingMode"></span><span id="_CPPv2NK12tensorrt_llm7runtime23SpeculativeDecodingModeeqERK23SpeculativeDecodingMode"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::eq-operator__SpeculativeDecodingModeCRC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1a545625bd71856b9ed609b9424ad09fef"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="k"><span class="pre">operator</span></span><span class="o"><span class="pre">==</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingModeE" title="tensorrt_llm::runtime::SpeculativeDecodingMode"><span class="n"><span class="pre">SpeculativeDecodingMode</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">other</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingModeeqERK23SpeculativeDecodingMode" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode23SpeculativeDecodingModeE14UnderlyingType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23SpeculativeDecodingMode23SpeculativeDecodingModeE14UnderlyingType"></span><span id="_CPPv2N12tensorrt_llm7runtime23SpeculativeDecodingMode23SpeculativeDecodingModeE14UnderlyingType"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::SpeculativeDecodingMode__UnderlyingTypeCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1a27fe9396ebb4470673dafa60eecf6db5"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">SpeculativeDecodingMode</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode14UnderlyingTypeE" title="tensorrt_llm::runtime::SpeculativeDecodingMode::UnderlyingType"><span class="n"><span class="pre">UnderlyingType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">state</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode23SpeculativeDecodingModeE14UnderlyingType" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-functions">Public Static Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode4NoneEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23SpeculativeDecodingMode4NoneEv"></span><span id="_CPPv2N12tensorrt_llm7runtime23SpeculativeDecodingMode4NoneEv"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::NoneCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1a0e993a88f1f7fd5b110459fb9aef8142"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">None</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode4NoneEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode19DraftTokensExternalEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23SpeculativeDecodingMode19DraftTokensExternalEv"></span><span id="_CPPv2N12tensorrt_llm7runtime23SpeculativeDecodingMode19DraftTokensExternalEv"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::DraftTokensExternalCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1a34e2a084be1ba4dc7f1fddba221bb9df"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">DraftTokensExternal</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode19DraftTokensExternalEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode6MedusaEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23SpeculativeDecodingMode6MedusaEv"></span><span id="_CPPv2N12tensorrt_llm7runtime23SpeculativeDecodingMode6MedusaEv"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::MedusaCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1a37730089e251e3ccd1e6e50d3ac2dead"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Medusa</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode6MedusaEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode17LookaheadDecodingEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23SpeculativeDecodingMode17LookaheadDecodingEv"></span><span id="_CPPv2N12tensorrt_llm7runtime23SpeculativeDecodingMode17LookaheadDecodingEv"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::LookaheadDecodingCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1a87b917526f7f2b0d821f7e034610649c"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">LookaheadDecoding</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode17LookaheadDecodingEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode19ExplicitDraftTokensEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23SpeculativeDecodingMode19ExplicitDraftTokensEv"></span><span id="_CPPv2N12tensorrt_llm7runtime23SpeculativeDecodingMode19ExplicitDraftTokensEv"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::ExplicitDraftTokensCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1ae7738d4022d1a16bbde026f7ae69acbf"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">ExplicitDraftTokens</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode19ExplicitDraftTokensEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode5EagleEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23SpeculativeDecodingMode5EagleEv"></span><span id="_CPPv2N12tensorrt_llm7runtime23SpeculativeDecodingMode5EagleEv"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::EagleCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1a18204f24dba152c9fc208659f7e97a3b"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Eagle</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode5EagleEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-functions">Private Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode9anyBitSetE14UnderlyingType">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime23SpeculativeDecodingMode9anyBitSetE14UnderlyingType"></span><span id="_CPPv2NK12tensorrt_llm7runtime23SpeculativeDecodingMode9anyBitSetE14UnderlyingType"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::anyBitSet__UnderlyingTypeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1a0a63f2f97f693e4c860330753711cdd6"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">anyBitSet</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode14UnderlyingTypeE" title="tensorrt_llm::runtime::SpeculativeDecodingMode::UnderlyingType"><span class="n"><span class="pre">UnderlyingType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">bits</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode9anyBitSetE14UnderlyingType" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode9allBitSetE14UnderlyingType">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime23SpeculativeDecodingMode9allBitSetE14UnderlyingType"></span><span id="_CPPv2NK12tensorrt_llm7runtime23SpeculativeDecodingMode9allBitSetE14UnderlyingType"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::allBitSet__UnderlyingTypeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1abc67823d81ebe2d45fbdbd7908e11153"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">allBitSet</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode14UnderlyingTypeE" title="tensorrt_llm::runtime::SpeculativeDecodingMode::UnderlyingType"><span class="n"><span class="pre">UnderlyingType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">bits</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime23SpeculativeDecodingMode9allBitSetE14UnderlyingType" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-members">Private Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode6mStateE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23SpeculativeDecodingMode6mStateE"></span><span id="_CPPv2N12tensorrt_llm7runtime23SpeculativeDecodingMode6mStateE"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::mState__UnderlyingType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1a0e8510f7bd29689984bc0cea9bff334f"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode14UnderlyingTypeE" title="tensorrt_llm::runtime::SpeculativeDecodingMode::UnderlyingType"><span class="n"><span class="pre">UnderlyingType</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mState</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode5kNoneE" title="tensorrt_llm::runtime::SpeculativeDecodingMode::kNone"><span class="n"><span class="pre">kNone</span></span></a><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode6mStateE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-static-attributes">Private Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode5kNoneE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23SpeculativeDecodingMode5kNoneE"></span><span id="_CPPv2N12tensorrt_llm7runtime23SpeculativeDecodingMode5kNoneE"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::kNone__UnderlyingType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1af78fd605d8b1f0fca8f5dbf4beb1618b"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode14UnderlyingTypeE" title="tensorrt_llm::runtime::SpeculativeDecodingMode::UnderlyingType"><span class="n"><span class="pre">UnderlyingType</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kNone</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">1U</span></span><span class="w"> </span><span class="o"><span class="pre"><<</span></span><span class="w"> </span><span class="m"><span class="pre">0U</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode5kNoneE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode20kDraftTokensExternalE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23SpeculativeDecodingMode20kDraftTokensExternalE"></span><span id="_CPPv2N12tensorrt_llm7runtime23SpeculativeDecodingMode20kDraftTokensExternalE"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::kDraftTokensExternal__UnderlyingType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1a096339698fc534bad97d16e3b044c461"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode14UnderlyingTypeE" title="tensorrt_llm::runtime::SpeculativeDecodingMode::UnderlyingType"><span class="n"><span class="pre">UnderlyingType</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kDraftTokensExternal</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">1U</span></span><span class="w"> </span><span class="o"><span class="pre"><<</span></span><span class="w"> </span><span class="m"><span class="pre">1U</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode20kDraftTokensExternalE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode7kMedusaE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23SpeculativeDecodingMode7kMedusaE"></span><span id="_CPPv2N12tensorrt_llm7runtime23SpeculativeDecodingMode7kMedusaE"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::kMedusa__UnderlyingType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1af315abd0c172f828f74b9e17abbd903f"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode14UnderlyingTypeE" title="tensorrt_llm::runtime::SpeculativeDecodingMode::UnderlyingType"><span class="n"><span class="pre">UnderlyingType</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kMedusa</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">1U</span></span><span class="w"> </span><span class="o"><span class="pre"><<</span></span><span class="w"> </span><span class="m"><span class="pre">2U</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode7kMedusaE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode18kLookaheadDecodingE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23SpeculativeDecodingMode18kLookaheadDecodingE"></span><span id="_CPPv2N12tensorrt_llm7runtime23SpeculativeDecodingMode18kLookaheadDecodingE"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::kLookaheadDecoding__UnderlyingType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1ae9223f63d456516b693af04eed4b1178"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode14UnderlyingTypeE" title="tensorrt_llm::runtime::SpeculativeDecodingMode::UnderlyingType"><span class="n"><span class="pre">UnderlyingType</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kLookaheadDecoding</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">1U</span></span><span class="w"> </span><span class="o"><span class="pre"><<</span></span><span class="w"> </span><span class="m"><span class="pre">3U</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode18kLookaheadDecodingE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode20kExplicitDraftTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23SpeculativeDecodingMode20kExplicitDraftTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime23SpeculativeDecodingMode20kExplicitDraftTokensE"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::kExplicitDraftTokens__UnderlyingType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1a7c8f3d5f099d84183f49969066c998da"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode14UnderlyingTypeE" title="tensorrt_llm::runtime::SpeculativeDecodingMode::UnderlyingType"><span class="n"><span class="pre">UnderlyingType</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kExplicitDraftTokens</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">1U</span></span><span class="w"> </span><span class="o"><span class="pre"><<</span></span><span class="w"> </span><span class="m"><span class="pre">4U</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode20kExplicitDraftTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode6kEagleE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23SpeculativeDecodingMode6kEagleE"></span><span id="_CPPv2N12tensorrt_llm7runtime23SpeculativeDecodingMode6kEagleE"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingMode::kEagle__UnderlyingType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingMode_1acf0f91631415d19f3b8cff019a1faf41"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode14UnderlyingTypeE" title="tensorrt_llm::runtime::SpeculativeDecodingMode::UnderlyingType"><span class="n"><span class="pre">UnderlyingType</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kEagle</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">1U</span></span><span class="w"> </span><span class="o"><span class="pre"><<</span></span><span class="w"> </span><span class="m"><span class="pre">5U</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingMode6kEagleE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="speculativedecodingmodule-h">
|
||
<h2>speculativeDecodingModule.h<a class="headerlink" href="#speculativedecodingmodule-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModuleE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime25SpeculativeDecodingModuleE"></span><span id="_CPPv2N12tensorrt_llm7runtime25SpeculativeDecodingModuleE"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingModule"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingModule"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">SpeculativeDecodingModule</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModuleE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Subclassed by <a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1LookaheadModule"><span class="std std-ref">tensorrt_llm::runtime::LookaheadModule</span></a>, <a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1MedusaModule"><span class="std std-ref">tensorrt_llm::runtime::MedusaModule</span></a></p>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule25SpeculativeDecodingModuleE10SizeType3210SizeType3210SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime25SpeculativeDecodingModule25SpeculativeDecodingModuleE10SizeType3210SizeType3210SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime25SpeculativeDecodingModule25SpeculativeDecodingModuleE10SizeType3210SizeType3210SizeType32"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingModule::SpeculativeDecodingModule__SizeType32.SizeType32.SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingModule_1aabf35face1ea9413d8b378bd70a33280"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">SpeculativeDecodingModule</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxDraftPathLen</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxDecodingDraftTokens</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxNumPaths</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule25SpeculativeDecodingModuleE10SizeType3210SizeType3210SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule25SpeculativeDecodingModuleEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime25SpeculativeDecodingModule25SpeculativeDecodingModuleEv"></span><span id="_CPPv2N12tensorrt_llm7runtime25SpeculativeDecodingModule25SpeculativeDecodingModuleEv"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingModule::SpeculativeDecodingModule"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingModule_1a1af69740c57c1f1a8e7590caf30b1a44"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">SpeculativeDecodingModule</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule25SpeculativeDecodingModuleEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModuleD0Ev">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime25SpeculativeDecodingModuleD0Ev"></span><span id="_CPPv2N12tensorrt_llm7runtime25SpeculativeDecodingModuleD0Ev"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingModule::~SpeculativeDecodingModule"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingModule_1a59776b8ad4a90d0a906bf00c619554da"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">~SpeculativeDecodingModule</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">default</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModuleD0Ev" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule25SpeculativeDecodingModuleERK25SpeculativeDecodingModule">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime25SpeculativeDecodingModule25SpeculativeDecodingModuleERK25SpeculativeDecodingModule"></span><span id="_CPPv2N12tensorrt_llm7runtime25SpeculativeDecodingModule25SpeculativeDecodingModuleERK25SpeculativeDecodingModule"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingModule::SpeculativeDecodingModule__SpeculativeDecodingModuleCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingModule_1a0793cd049c1b7502c683ad8c8da48008"></span><span class="sig-name descname"><span class="n"><span class="pre">SpeculativeDecodingModule</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule25SpeculativeDecodingModuleERK25SpeculativeDecodingModule" title="tensorrt_llm::runtime::SpeculativeDecodingModule::SpeculativeDecodingModule"><span class="n"><span class="pre">SpeculativeDecodingModule</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">o</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">default</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule25SpeculativeDecodingModuleERK25SpeculativeDecodingModule" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModuleaSERK25SpeculativeDecodingModule">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime25SpeculativeDecodingModuleaSERK25SpeculativeDecodingModule"></span><span id="_CPPv2N12tensorrt_llm7runtime25SpeculativeDecodingModuleaSERK25SpeculativeDecodingModule"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingModule::assign-operator__SpeculativeDecodingModuleCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingModule_1ad166641565c8f32ebf6afdc788b546a3"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModuleE" title="tensorrt_llm::runtime::SpeculativeDecodingModule"><span class="n"><span class="pre">SpeculativeDecodingModule</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="k"><span class="pre">operator</span></span><span class="o"><span class="pre">=</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModuleE" title="tensorrt_llm::runtime::SpeculativeDecodingModule"><span class="n"><span class="pre">SpeculativeDecodingModule</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">o</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">default</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModuleaSERK25SpeculativeDecodingModule" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime25SpeculativeDecodingModule18getMaxDraftPathLenEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime25SpeculativeDecodingModule18getMaxDraftPathLenEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime25SpeculativeDecodingModule18getMaxDraftPathLenEv"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingModule::getMaxDraftPathLenC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingModule_1a92a91b11ffd70a0fd2698cd6f1c96a69"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getMaxDraftPathLen</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime25SpeculativeDecodingModule18getMaxDraftPathLenEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>max number of draft tokens that can be accepted by one step of the decoder </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime25SpeculativeDecodingModule13getMaxPathLenEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime25SpeculativeDecodingModule13getMaxPathLenEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime25SpeculativeDecodingModule13getMaxPathLenEv"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingModule::getMaxPathLenC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingModule_1ab4b8c33566b684613b230b0b740e0ded"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getMaxPathLen</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime25SpeculativeDecodingModule13getMaxPathLenEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>
|
||
one more than draft path len for prediction from primary head </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>max number of tokens that a request can grow in one step of the decoder</p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime25SpeculativeDecodingModule25getMaxDecodingDraftTokensEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime25SpeculativeDecodingModule25getMaxDecodingDraftTokensEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime25SpeculativeDecodingModule25getMaxDecodingDraftTokensEv"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingModule::getMaxDecodingDraftTokensC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingModule_1a3b693afe821b966e8312236e16f527f2"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getMaxDecodingDraftTokens</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime25SpeculativeDecodingModule25getMaxDecodingDraftTokensEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>max number of draft tokens processed by one step of the decoder </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime25SpeculativeDecodingModule20getMaxDecodingTokensEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime25SpeculativeDecodingModule20getMaxDecodingTokensEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime25SpeculativeDecodingModule20getMaxDecodingTokensEv"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingModule::getMaxDecodingTokensC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingModule_1a6fe241359c3e7969a7b506f4a9b431e3"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getMaxDecodingTokens</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime25SpeculativeDecodingModule20getMaxDecodingTokensEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>
|
||
one more than decoding draft tokens for prediction from primary head </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>max number of tokens processed by one step of the decoder</p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime25SpeculativeDecodingModule17getNumPackedMasksEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime25SpeculativeDecodingModule17getNumPackedMasksEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime25SpeculativeDecodingModule17getNumPackedMasksEv"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingModule::getNumPackedMasksC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingModule_1a78cb7cbc32f4f2cb9a2864c65c3c0d0a"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getNumPackedMasks</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime25SpeculativeDecodingModule17getNumPackedMasksEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime25SpeculativeDecodingModule14getMaxNumPathsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime25SpeculativeDecodingModule14getMaxNumPathsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime25SpeculativeDecodingModule14getMaxNumPathsEv"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingModule::getMaxNumPathsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingModule_1a462630e52247e02e3c01c98a23d1f705"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getMaxNumPaths</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime25SpeculativeDecodingModule14getMaxNumPathsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule17setMaxDraftTokensE10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime25SpeculativeDecodingModule17setMaxDraftTokensE10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime25SpeculativeDecodingModule17setMaxDraftTokensE10SizeType32"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingModule::setMaxDraftTokens__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingModule_1a6bfb797bb4ba7bfba00fbbdec168e983"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setMaxDraftTokens</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxDraftTokens</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule17setMaxDraftTokensE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule18setMaxDraftPathLenE10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime25SpeculativeDecodingModule18setMaxDraftPathLenE10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime25SpeculativeDecodingModule18setMaxDraftPathLenE10SizeType32"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingModule::setMaxDraftPathLen__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingModule_1a6f550a2b2d37b70fd653b5738da40be3"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setMaxDraftPathLen</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxDraftPathLen</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule18setMaxDraftPathLenE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule14setMaxNumPathsE10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime25SpeculativeDecodingModule14setMaxNumPathsE10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime25SpeculativeDecodingModule14setMaxNumPathsE10SizeType32"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingModule::setMaxNumPaths__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingModule_1a84e469ace7d51c3fe1d38d02a679480f"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setMaxNumPaths</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxNumPaths</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule14setMaxNumPathsE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-functions">Private Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule21computeNumPackedMasksEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime25SpeculativeDecodingModule21computeNumPackedMasksEv"></span><span id="_CPPv2N12tensorrt_llm7runtime25SpeculativeDecodingModule21computeNumPackedMasksEv"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingModule::computeNumPackedMasks"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingModule_1af7f830a6266ee46c0b91414b70c36c6b"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">computeNumPackedMasks</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule21computeNumPackedMasksEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-members">Private Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule16mMaxDraftPathLenE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime25SpeculativeDecodingModule16mMaxDraftPathLenE"></span><span id="_CPPv2N12tensorrt_llm7runtime25SpeculativeDecodingModule16mMaxDraftPathLenE"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingModule::mMaxDraftPathLen__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingModule_1a8bf6280178a97c5e534ac182bb4f301f"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mMaxDraftPathLen</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule16mMaxDraftPathLenE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule23mMaxDecodingDraftTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime25SpeculativeDecodingModule23mMaxDecodingDraftTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime25SpeculativeDecodingModule23mMaxDecodingDraftTokensE"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingModule::mMaxDecodingDraftTokens__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingModule_1af863071adbd70ecffa6659a361c200de"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mMaxDecodingDraftTokens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule23mMaxDecodingDraftTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule12mMaxNumPathsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime25SpeculativeDecodingModule12mMaxNumPathsE"></span><span id="_CPPv2N12tensorrt_llm7runtime25SpeculativeDecodingModule12mMaxNumPathsE"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingModule::mMaxNumPaths__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingModule_1a00d82632302e77fd010e983272e65410"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mMaxNumPaths</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule12mMaxNumPathsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule18mMaxNumPackedMasksE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime25SpeculativeDecodingModule18mMaxNumPackedMasksE"></span><span id="_CPPv2N12tensorrt_llm7runtime25SpeculativeDecodingModule18mMaxNumPackedMasksE"></span><span id="tensorrt_llm::runtime::SpeculativeDecodingModule::mMaxNumPackedMasks__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SpeculativeDecodingModule_1a0229e91e73bbb60aa933cc5c37e1deab"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mMaxNumPackedMasks</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModule18mMaxNumPackedMasksE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="tllmlogger-h">
|
||
<h2>tllmLogger.h<a class="headerlink" href="#tllmlogger-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10TllmLoggerE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10TllmLoggerE"></span><span id="_CPPv2N12tensorrt_llm7runtime10TllmLoggerE"></span><span id="tensorrt_llm::runtime::TllmLogger"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1TllmLogger"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TllmLogger</span></span></span><span class="w"> </span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="k"><span class="pre">public</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ILogger</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10TllmLoggerE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10TllmLogger3logE8SeverityPKN8nvinfer19AsciiCharE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10TllmLogger3logE8SeverityPKN8nvinfer19AsciiCharE"></span><span id="_CPPv2N12tensorrt_llm7runtime10TllmLogger3logE8SeverityPKN8nvinfer19AsciiCharE"></span><span id="tensorrt_llm::runtime::TllmLogger::log__Severity.nvinfer1::AsciiCharCP"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1TllmLogger_1a052c775ee14bc0d741d26d28c5b3f311"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">log</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">Severity</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">severity</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">AsciiChar</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="n sig-param"><span class="pre">msg</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10TllmLogger3logE8SeverityPKN8nvinfer19AsciiCharE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10TllmLogger8getLevelEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10TllmLogger8getLevelEv"></span><span id="_CPPv2N12tensorrt_llm7runtime10TllmLogger8getLevelEv"></span><span id="tensorrt_llm::runtime::TllmLogger::getLevel"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1TllmLogger_1ae6b565ac5ee12cfcd305c0f2c0bd4b1e"></span><span class="n"><span class="pre">Severity</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getLevel</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10TllmLogger8getLevelEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10TllmLogger8setLevelE8Severity">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10TllmLogger8setLevelE8Severity"></span><span id="_CPPv2N12tensorrt_llm7runtime10TllmLogger8setLevelE8Severity"></span><span id="tensorrt_llm::runtime::TllmLogger::setLevel__Severity"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1TllmLogger_1a02ca7ebe0eec266f8b6ab4b66e9f0275"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setLevel</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">Severity</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">level</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10TllmLogger8setLevelE8Severity" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="worldconfig-h">
|
||
<h2>worldConfig.h<a class="headerlink" href="#worldconfig-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11WorldConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11WorldConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime11WorldConfigE"></span><span id="tensorrt_llm::runtime::WorldConfig"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">WorldConfig</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11WorldConfig11WorldConfigE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType32RKNSt8optionalINSt6vectorI10SizeType32EEEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11WorldConfig11WorldConfigE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType32RKNSt8optionalINSt6vectorI10SizeType32EEEE"></span><span id="_CPPv2N12tensorrt_llm7runtime11WorldConfig11WorldConfigE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType32RKNSt8optionalINSt6vectorI10SizeType32EEEE"></span><span id="tensorrt_llm::runtime::WorldConfig::WorldConfig__SizeType32.SizeType32.SizeType32.SizeType32.SizeType32.std::optional:std::vector:SizeType32::CR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1aa383b7829757256b4a18a19268b41d76"></span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">WorldConfig</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tensorParallelism</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">1</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">pipelineParallelism</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">1</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">contextParallelism</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">1</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">rank</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">gpusPerNode</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfig19kDefaultGpusPerNodeE" title="tensorrt_llm::runtime::WorldConfig::kDefaultGpusPerNode"><span class="n"><span class="pre">kDefaultGpusPerNode</span></span></a>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">deviceIds</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">nullopt</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfig11WorldConfigE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType32RKNSt8optionalINSt6vectorI10SizeType32EEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11WorldConfig7getSizeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11WorldConfig7getSizeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11WorldConfig7getSizeEv"></span><span id="tensorrt_llm::runtime::WorldConfig::getSizeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1a8b9d102af9689ecca7b6e9924ca955a2"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getSize</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig7getSizeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11WorldConfig20getTensorParallelismEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11WorldConfig20getTensorParallelismEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11WorldConfig20getTensorParallelismEv"></span><span id="tensorrt_llm::runtime::WorldConfig::getTensorParallelismCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1ae5022fa448a9d76e460b1a255d47c9e3"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getTensorParallelism</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig20getTensorParallelismEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11WorldConfig16isTensorParallelEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11WorldConfig16isTensorParallelEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11WorldConfig16isTensorParallelEv"></span><span id="tensorrt_llm::runtime::WorldConfig::isTensorParallelCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1a00a72e2f83f447679b12024100e2bd51"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">isTensorParallel</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig16isTensorParallelEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11WorldConfig22getPipelineParallelismEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11WorldConfig22getPipelineParallelismEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11WorldConfig22getPipelineParallelismEv"></span><span id="tensorrt_llm::runtime::WorldConfig::getPipelineParallelismCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1ac69e3f6afd55e830b76b6a39a14481cd"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getPipelineParallelism</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig22getPipelineParallelismEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11WorldConfig18isPipelineParallelEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11WorldConfig18isPipelineParallelEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11WorldConfig18isPipelineParallelEv"></span><span id="tensorrt_llm::runtime::WorldConfig::isPipelineParallelCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1a26eaecd483304e8df407068905d9123c"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">isPipelineParallel</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig18isPipelineParallelEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11WorldConfig21getContextParallelismEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11WorldConfig21getContextParallelismEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11WorldConfig21getContextParallelismEv"></span><span id="tensorrt_llm::runtime::WorldConfig::getContextParallelismCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1a004b171a1af6b36bcb45df247c77485a"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getContextParallelism</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig21getContextParallelismEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11WorldConfig17isContextParallelEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11WorldConfig17isContextParallelEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11WorldConfig17isContextParallelEv"></span><span id="tensorrt_llm::runtime::WorldConfig::isContextParallelCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1a7258b051cb4ac27b5a99e5999467c733"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">isContextParallel</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig17isContextParallelEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11WorldConfig7getRankEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11WorldConfig7getRankEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11WorldConfig7getRankEv"></span><span id="tensorrt_llm::runtime::WorldConfig::getRankCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1af1cb5b83608c3da757e7dbe2b1e5597a"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getRank</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig7getRankEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11WorldConfig14getGpusPerNodeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11WorldConfig14getGpusPerNodeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11WorldConfig14getGpusPerNodeEv"></span><span id="tensorrt_llm::runtime::WorldConfig::getGpusPerNodeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1a44a3e7694a2c357f5b5d63e5964cfcb2"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getGpusPerNode</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig14getGpusPerNodeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11WorldConfig15getGpusPerGroupEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11WorldConfig15getGpusPerGroupEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11WorldConfig15getGpusPerGroupEv"></span><span id="tensorrt_llm::runtime::WorldConfig::getGpusPerGroupC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1adcd82f3f12d0fa200af350aa7e6c03fc"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getGpusPerGroup</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig15getGpusPerGroupEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11WorldConfig9getDeviceEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11WorldConfig9getDeviceEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11WorldConfig9getDeviceEv"></span><span id="tensorrt_llm::runtime::WorldConfig::getDeviceC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1a0e1c32dce89cf5bb8a0c6442254b77aa"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getDevice</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig9getDeviceEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11WorldConfig11getDeviceOfE10SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11WorldConfig11getDeviceOfE10SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime11WorldConfig11getDeviceOfE10SizeType32"></span><span id="tensorrt_llm::runtime::WorldConfig::getDeviceOf__SizeType32C"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1aa10799e03062dbc43bba2c25136ebf74"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getDeviceOf</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">rank</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig11getDeviceOfE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11WorldConfig23getPipelineParallelRankEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11WorldConfig23getPipelineParallelRankEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11WorldConfig23getPipelineParallelRankEv"></span><span id="tensorrt_llm::runtime::WorldConfig::getPipelineParallelRankCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1a4edd655c3bd2758d67f0171d77e54f5d"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getPipelineParallelRank</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig23getPipelineParallelRankEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11WorldConfig21getTensorParallelRankEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11WorldConfig21getTensorParallelRankEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11WorldConfig21getTensorParallelRankEv"></span><span id="tensorrt_llm::runtime::WorldConfig::getTensorParallelRankCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1a63c6d87c37aadcd07700dd935b4a91e5"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getTensorParallelRank</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig21getTensorParallelRankEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11WorldConfig22getContextParallelRankEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11WorldConfig22getContextParallelRankEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11WorldConfig22getContextParallelRankEv"></span><span id="tensorrt_llm::runtime::WorldConfig::getContextParallelRankCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1aa257d2b25d50ee74832f93c179b9ee41"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getContextParallelRank</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig22getContextParallelRankEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11WorldConfig12getLocalRankEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11WorldConfig12getLocalRankEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11WorldConfig12getLocalRankEv"></span><span id="tensorrt_llm::runtime::WorldConfig::getLocalRankCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1af7740d5c7afd1ba7a98f4b2e0f481838"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getLocalRank</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig12getLocalRankEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11WorldConfig11getNodeRankEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11WorldConfig11getNodeRankEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11WorldConfig11getNodeRankEv"></span><span id="tensorrt_llm::runtime::WorldConfig::getNodeRankCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1a4e5ed903b009aee7f656931e4902c8ce"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getNodeRank</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig11getNodeRankEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11WorldConfig13getNodeRankOfE10SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11WorldConfig13getNodeRankOfE10SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime11WorldConfig13getNodeRankOfE10SizeType32"></span><span id="tensorrt_llm::runtime::WorldConfig::getNodeRankOf__SizeType32CCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1ad0c5b7241276daca31e02c9305ea7fa2"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getNodeRankOf</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">rank</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig13getNodeRankOfE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11WorldConfig27isFirstPipelineParallelRankEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11WorldConfig27isFirstPipelineParallelRankEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11WorldConfig27isFirstPipelineParallelRankEv"></span><span id="tensorrt_llm::runtime::WorldConfig::isFirstPipelineParallelRankCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1a23a7f6bb812b6d0e60325e91c14cb2e0"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">isFirstPipelineParallelRank</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig27isFirstPipelineParallelRankEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11WorldConfig26isLastPipelineParallelRankEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11WorldConfig26isLastPipelineParallelRankEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11WorldConfig26isLastPipelineParallelRankEv"></span><span id="tensorrt_llm::runtime::WorldConfig::isLastPipelineParallelRankCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1a86b29f5ea72282f86f8af979edb6c3e7"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">isLastPipelineParallelRank</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig26isLastPipelineParallelRankEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Is my rank the last rank in its pipeline? </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11WorldConfig25isFirstTensorParallelRankEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11WorldConfig25isFirstTensorParallelRankEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11WorldConfig25isFirstTensorParallelRankEv"></span><span id="tensorrt_llm::runtime::WorldConfig::isFirstTensorParallelRankCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1a00d09d798d8301bb87dd364f7a47193f"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">isFirstTensorParallelRank</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig25isFirstTensorParallelRankEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11WorldConfig26isFirstContextParallelRankEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11WorldConfig26isFirstContextParallelRankEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11WorldConfig26isFirstContextParallelRankEv"></span><span id="tensorrt_llm::runtime::WorldConfig::isFirstContextParallelRankCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1a1134cb4738755b321c00e886ab716ac2"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">isFirstContextParallelRank</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig26isFirstContextParallelRankEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11WorldConfig11getLastRankEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11WorldConfig11getLastRankEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11WorldConfig11getLastRankEv"></span><span id="tensorrt_llm::runtime::WorldConfig::getLastRankCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1a8cf59e9a62e4801a2ae25f3b0cbc2e89"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getLastRank</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig11getLastRankEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11WorldConfig24getPipelineParallelGroupEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11WorldConfig24getPipelineParallelGroupEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11WorldConfig24getPipelineParallelGroupEv"></span><span id="tensorrt_llm::runtime::WorldConfig::getPipelineParallelGroupC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1ad3ce7871a43bf0bf6ca72346b6605e02"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getPipelineParallelGroup</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig24getPipelineParallelGroupEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11WorldConfig22getTensorParallelGroupEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11WorldConfig22getTensorParallelGroupEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11WorldConfig22getTensorParallelGroupEv"></span><span id="tensorrt_llm::runtime::WorldConfig::getTensorParallelGroupC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1aac48fc5479138ba138634326ef49ed01"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getTensorParallelGroup</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig22getTensorParallelGroupEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11WorldConfig23getContextParallelGroupEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11WorldConfig23getContextParallelGroupEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11WorldConfig23getContextParallelGroupEv"></span><span id="tensorrt_llm::runtime::WorldConfig::getContextParallelGroupC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1afbd5f464ff91ee9ca154b6c7d3b5447a"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getContextParallelGroup</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig23getContextParallelGroupEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11WorldConfig14validMpiConfigEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11WorldConfig14validMpiConfigEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11WorldConfig14validMpiConfigEv"></span><span id="tensorrt_llm::runtime::WorldConfig::validMpiConfigC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1a92e7fa800262ca7d7ca08f2705d30626"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">validMpiConfig</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11WorldConfig14validMpiConfigEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-functions">Public Static Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11WorldConfig3mpiE10SizeType32NSt8optionalI10SizeType32EENSt8optionalI10SizeType32EENSt8optionalI10SizeType32EERKNSt8optionalINSt6vectorI10SizeType32EEEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11WorldConfig3mpiE10SizeType32NSt8optionalI10SizeType32EENSt8optionalI10SizeType32EENSt8optionalI10SizeType32EERKNSt8optionalINSt6vectorI10SizeType32EEEE"></span><span id="_CPPv2N12tensorrt_llm7runtime11WorldConfig3mpiE10SizeType32NSt8optionalI10SizeType32EENSt8optionalI10SizeType32EENSt8optionalI10SizeType32EERKNSt8optionalINSt6vectorI10SizeType32EEEE"></span><span id="tensorrt_llm::runtime::WorldConfig::mpi__SizeType32.std::optional:SizeType32:.std::optional:SizeType32:.std::optional:SizeType32:.std::optional:std::vector:SizeType32::CR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1aa633c302a519dbae62201b2f082d5675"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfigE" title="tensorrt_llm::runtime::WorldConfig"><span class="n"><span class="pre">WorldConfig</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mpi</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">gpusPerNode</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfig19kDefaultGpusPerNodeE" title="tensorrt_llm::runtime::WorldConfig::kDefaultGpusPerNode"><span class="n"><span class="pre">kDefaultGpusPerNode</span></span></a>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="n sig-param"><span class="pre">tensorParallelism</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">nullopt</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="n sig-param"><span class="pre">pipelineParallelism</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">nullopt</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="n sig-param"><span class="pre">contextParallelism</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">nullopt</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">deviceIds</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">nullopt</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfig3mpiE10SizeType32NSt8optionalI10SizeType32EENSt8optionalI10SizeType32EENSt8optionalI10SizeType32EERKNSt8optionalINSt6vectorI10SizeType32EEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11WorldConfig19kDefaultGpusPerNodeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11WorldConfig19kDefaultGpusPerNodeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11WorldConfig19kDefaultGpusPerNodeE"></span><span id="tensorrt_llm::runtime::WorldConfig::kDefaultGpusPerNode__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1a947b944a0ba919cf264b2f40d6e88fe1"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kDefaultGpusPerNode</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">1</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfig19kDefaultGpusPerNodeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-members">Private Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11WorldConfig18mTensorParallelismE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11WorldConfig18mTensorParallelismE"></span><span id="_CPPv2N12tensorrt_llm7runtime11WorldConfig18mTensorParallelismE"></span><span id="tensorrt_llm::runtime::WorldConfig::mTensorParallelism__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1aa7b502999329a2e6c0befbec8bb391d7"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mTensorParallelism</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfig18mTensorParallelismE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11WorldConfig20mPipelineParallelismE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11WorldConfig20mPipelineParallelismE"></span><span id="_CPPv2N12tensorrt_llm7runtime11WorldConfig20mPipelineParallelismE"></span><span id="tensorrt_llm::runtime::WorldConfig::mPipelineParallelism__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1af25b064d4e53a41f5c73ad2c2e7798b9"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mPipelineParallelism</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfig20mPipelineParallelismE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11WorldConfig19mContextParallelismE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11WorldConfig19mContextParallelismE"></span><span id="_CPPv2N12tensorrt_llm7runtime11WorldConfig19mContextParallelismE"></span><span id="tensorrt_llm::runtime::WorldConfig::mContextParallelism__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1aac24e266668057de079b5cf50d9df978"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mContextParallelism</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfig19mContextParallelismE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11WorldConfig5mRankE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11WorldConfig5mRankE"></span><span id="_CPPv2N12tensorrt_llm7runtime11WorldConfig5mRankE"></span><span id="tensorrt_llm::runtime::WorldConfig::mRank__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1a5f0fde85e5fe37245b4f8e544910dd29"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mRank</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfig5mRankE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11WorldConfig12mGpusPerNodeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11WorldConfig12mGpusPerNodeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11WorldConfig12mGpusPerNodeE"></span><span id="tensorrt_llm::runtime::WorldConfig::mGpusPerNode__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1a2db2a3ff84174617be9b1de7833f6792"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mGpusPerNode</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfig12mGpusPerNodeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11WorldConfig10mDeviceIdsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11WorldConfig10mDeviceIdsE"></span><span id="_CPPv2N12tensorrt_llm7runtime11WorldConfig10mDeviceIdsE"></span><span id="tensorrt_llm::runtime::WorldConfig::mDeviceIds__std::vector:SizeType32:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1a4e6848ca14ea58630295ffb14c365e39"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mDeviceIds</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfig10mDeviceIdsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="eaglebuffers-h">
|
||
<h2>eagleBuffers.h<a class="headerlink" href="#eaglebuffers-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1batch__manager"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">batch_manager</span></span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffersE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffersE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffersE"></span><span id="tensorrt_llm::runtime::EagleBuffers"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">EagleBuffers</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffersE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers13LlmRequestPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers13LlmRequestPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers13LlmRequestPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1a69aa6170271f65247462172a15600c88"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">LlmRequestPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm13batch_managerE" title="tensorrt_llm::batch_manager"><span class="n"><span class="pre">batch_manager</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">LlmRequest</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers13LlmRequestPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers13RequestVectorE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers13RequestVectorE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers13RequestVectorE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1a36d74a637a8d68ab93d55e9af634471a"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">RequestVector</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers13LlmRequestPtrE" title="tensorrt_llm::runtime::EagleBuffers::LlmRequestPtr"><span class="n"><span class="pre">LlmRequestPtr</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers13RequestVectorE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers10SizeType32E">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers10SizeType32E"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers10SizeType32E"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1a3626c46ed5783f220200077cd9fee59f"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">SizeType32</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers10SizeType32E" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers7ITensorE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers7ITensorE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers7ITensorE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1a73d1859dd18dab0ef46c990054dc2327"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">ITensor</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers7ITensorE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers9BufferPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers9BufferPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers9BufferPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1adc9567ee5765b159fc4fc7bc38251eee"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">BufferPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer9SharedPtrE" title="tensorrt_llm::runtime::IBuffer::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9BufferPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1ae97d295f67ee450da4d4c512daa21413"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorMapE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers9TensorMapE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers9TensorMapE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1ac84f2deab51cdf8ddc998aaf4cf96e18"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorMap</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime12StringPtrMapE" title="tensorrt_llm::runtime::StringPtrMap"><span class="n"><span class="pre">StringPtrMap</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorMapE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers12EagleBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerERKN7runtime11ModelConfigERKN7runtime11WorldConfigERKN8executor14DecodingConfigERKN7runtime11TllmRuntimeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers12EagleBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerERKN7runtime11ModelConfigERKN7runtime11WorldConfigERKN8executor14DecodingConfigERKN7runtime11TllmRuntimeE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers12EagleBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerERKN7runtime11ModelConfigERKN7runtime11WorldConfigERKN8executor14DecodingConfigERKN7runtime11TllmRuntimeE"></span><span id="tensorrt_llm::runtime::EagleBuffers::EagleBuffers__SizeType32.SizeType32.runtime::BufferManagerCR.runtime::ModelConfigCR.runtime::WorldConfigCR.executor::DecodingConfigCR.runtime::TllmRuntimeCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1a72967b030b6b3862ca6f60e78c429e49"></span><span class="sig-name descname"><span class="n"><span class="pre">EagleBuffers</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers10SizeType32E" title="tensorrt_llm::runtime::EagleBuffers::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxBatchSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers10SizeType32E" title="tensorrt_llm::runtime::EagleBuffers::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxBeamWidth</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers12EagleBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerERKN7runtime11ModelConfigERKN7runtime11WorldConfigERKN8executor14DecodingConfigERKN7runtime11TllmRuntimeE" title="tensorrt_llm::runtime::EagleBuffers::EagleBuffers::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">BufferManager</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">manager</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers12EagleBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerERKN7runtime11ModelConfigERKN7runtime11WorldConfigERKN8executor14DecodingConfigERKN7runtime11TllmRuntimeE" title="tensorrt_llm::runtime::EagleBuffers::EagleBuffers::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ModelConfig</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers12EagleBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerERKN7runtime11ModelConfigERKN7runtime11WorldConfigERKN8executor14DecodingConfigERKN7runtime11TllmRuntimeE" title="tensorrt_llm::runtime::EagleBuffers::EagleBuffers::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">WorldConfig</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">worldConfig</span></span>, <a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executorE" title="tensorrt_llm::executor"><span class="n"><span class="pre">executor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executor14DecodingConfigE" title="tensorrt_llm::executor::DecodingConfig"><span class="n"><span class="pre">DecodingConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">decodingConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers12EagleBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerERKN7runtime11ModelConfigERKN7runtime11WorldConfigERKN8executor14DecodingConfigERKN7runtime11TllmRuntimeE" title="tensorrt_llm::runtime::EagleBuffers::EagleBuffers::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">TllmRuntime</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">runtime</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers12EagleBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerERKN7runtime11ModelConfigERKN7runtime11WorldConfigERKN8executor14DecodingConfigERKN7runtime11TllmRuntimeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers7reshapeE10SizeType3210SizeType32RKN7runtime11ModelConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers7reshapeE10SizeType3210SizeType32RKN7runtime11ModelConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers7reshapeE10SizeType3210SizeType32RKN7runtime11ModelConfigE"></span><span id="tensorrt_llm::runtime::EagleBuffers::reshape__SizeType32.SizeType32.runtime::ModelConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1ab7d4d9270143f5aa163a5c66b52e9f2a"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">reshape</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers10SizeType32E" title="tensorrt_llm::runtime::EagleBuffers::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">numCtxSequences</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers10SizeType32E" title="tensorrt_llm::runtime::EagleBuffers::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">numGenSequences</span></span>, <a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE" title="tensorrt_llm::runtime::ModelConfig"><span class="n"><span class="pre">ModelConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelConfig</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers7reshapeE10SizeType3210SizeType32RKN7runtime11ModelConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime12EagleBuffers13setFromInputsERK13RequestVectorRK13RequestVectorRKN7runtime7ITensorERK7ITensorRKN12EagleBuffers6InputsERKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime12EagleBuffers13setFromInputsERK13RequestVectorRK13RequestVectorRKN7runtime7ITensorERK7ITensorRKN12EagleBuffers6InputsERKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE"></span><span id="_CPPv2NK12tensorrt_llm7runtime12EagleBuffers13setFromInputsERK13RequestVectorRK13RequestVectorRKN7runtime7ITensorERK7ITensorRKN12EagleBuffers6InputsERKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE"></span><span id="tensorrt_llm::runtime::EagleBuffers::setFromInputs__RequestVectorCR.RequestVectorCR.runtime::ITensorCR.ITensorCR.EagleBuffers::InputsCR.runtime::TllmRuntimeCR.runtime::ModelConfigCR.runtime::WorldConfigCRC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1ae8c4b732c3f1942501dfccb6aaad7300"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setFromInputs</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers13RequestVectorE" title="tensorrt_llm::runtime::EagleBuffers::RequestVector"><span class="n"><span class="pre">RequestVector</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">contextRequests</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers13RequestVectorE" title="tensorrt_llm::runtime::EagleBuffers::RequestVector"><span class="n"><span class="pre">RequestVector</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">genRequests</span></span>, <a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime12EagleBuffers13setFromInputsERK13RequestVectorRK13RequestVectorRKN7runtime7ITensorERK7ITensorRKN12EagleBuffers6InputsERKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE" title="tensorrt_llm::runtime::EagleBuffers::setFromInputs::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ITensor</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">requestTypes</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers7ITensorE" title="tensorrt_llm::runtime::EagleBuffers::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">seqSlots</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffersE" title="tensorrt_llm::runtime::EagleBuffers"><span class="n"><span class="pre">EagleBuffers</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6InputsE" title="tensorrt_llm::runtime::EagleBuffers::Inputs"><span class="n"><span class="pre">Inputs</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">decoderBuffers</span></span>, <a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime12EagleBuffers13setFromInputsERK13RequestVectorRK13RequestVectorRKN7runtime7ITensorERK7ITensorRKN12EagleBuffers6InputsERKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE" title="tensorrt_llm::runtime::EagleBuffers::setFromInputs::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">TllmRuntime</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">runtime</span></span>, <a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime12EagleBuffers13setFromInputsERK13RequestVectorRK13RequestVectorRKN7runtime7ITensorERK7ITensorRKN12EagleBuffers6InputsERKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE" title="tensorrt_llm::runtime::EagleBuffers::setFromInputs::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ModelConfig</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelConfig</span></span>, <a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime12EagleBuffers13setFromInputsERK13RequestVectorRK13RequestVectorRKN7runtime7ITensorERK7ITensorRKN12EagleBuffers6InputsERKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE" title="tensorrt_llm::runtime::EagleBuffers::setFromInputs::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">WorldConfig</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">worldConfig</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime12EagleBuffers13setFromInputsERK13RequestVectorRK13RequestVectorRKN7runtime7ITensorERK7ITensorRKN12EagleBuffers6InputsERKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime12EagleBuffers18insertInputTensorsER9TensorMapR9TensorMapRKN7runtime11WorldConfigE">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime12EagleBuffers18insertInputTensorsER9TensorMapR9TensorMapRKN7runtime11WorldConfigE"></span><span id="_CPPv2NK12tensorrt_llm7runtime12EagleBuffers18insertInputTensorsER9TensorMapR9TensorMapRKN7runtime11WorldConfigE"></span><span id="tensorrt_llm::runtime::EagleBuffers::insertInputTensors__TensorMapR.TensorMapR.runtime::WorldConfigCRC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1a82c9e53cd86889744a3383b89dc33f23"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">insertInputTensors</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorMapE" title="tensorrt_llm::runtime::EagleBuffers::TensorMap"><span class="n"><span class="pre">TensorMap</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">inputBuffers</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorMapE" title="tensorrt_llm::runtime::EagleBuffers::TensorMap"><span class="n"><span class="pre">TensorMap</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">outputBuffers</span></span>, <a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfigE" title="tensorrt_llm::runtime::WorldConfig"><span class="n"><span class="pre">WorldConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">worldConfig</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime12EagleBuffers18insertInputTensorsER9TensorMapR9TensorMapRKN7runtime11WorldConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers12engineInputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers12engineInputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers12engineInputsE"></span><span id="tensorrt_llm::runtime::EagleBuffers::engineInputs__Inputs"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1a9659dbf9f9d1e45eddcc5c3054b680dd"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6InputsE" title="tensorrt_llm::runtime::EagleBuffers::Inputs"><span class="n"><span class="pre">Inputs</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">engineInputs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers12engineInputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers13engineOutputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers13engineOutputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers13engineOutputsE"></span><span id="tensorrt_llm::runtime::EagleBuffers::engineOutputs__tensorrt_llm::runtime::EagleBuffers::EngineOutputs"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1abd3e4bcd264d87334f064ceef1afd60b"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffersE" title="tensorrt_llm::runtime::EagleBuffers"><span class="n"><span class="pre">EagleBuffers</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers13EngineOutputsE" title="tensorrt_llm::runtime::EagleBuffers::EngineOutputs"><span class="n"><span class="pre">EngineOutputs</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">engineOutputs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers13engineOutputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-functions">Private Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0ENK12tensorrt_llm7runtime12EagleBuffers13setFromInputsEvRK13RequestVectorRK13RequestVector10SizeType32RK7ITensorRKN12EagleBuffers6InputsERKN7runtime11EagleModuleERKN7runtime13BufferManagerE">
|
||
<span id="_CPPv3I0ENK12tensorrt_llm7runtime12EagleBuffers13setFromInputsERK13RequestVectorRK13RequestVector10SizeType32RK7ITensorRKN12EagleBuffers6InputsERKN7runtime11EagleModuleERKN7runtime13BufferManagerE"></span><span id="_CPPv2I0ENK12tensorrt_llm7runtime12EagleBuffers13setFromInputsERK13RequestVectorRK13RequestVector10SizeType32RK7ITensorRKN12EagleBuffers6InputsERKN7runtime11EagleModuleERKN7runtime13BufferManagerE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1a3ad1ef640dbd717e717a1d21c1f8a501"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setFromInputs</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers13RequestVectorE" title="tensorrt_llm::runtime::EagleBuffers::RequestVector"><span class="n"><span class="pre">RequestVector</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">contextRequests</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers13RequestVectorE" title="tensorrt_llm::runtime::EagleBuffers::RequestVector"><span class="n"><span class="pre">RequestVector</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">genRequests</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers10SizeType32E" title="tensorrt_llm::runtime::EagleBuffers::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">vocabSizePadded</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers7ITensorE" title="tensorrt_llm::runtime::EagleBuffers::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">seqSlots</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffersE" title="tensorrt_llm::runtime::EagleBuffers"><span class="n"><span class="pre">EagleBuffers</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6InputsE" title="tensorrt_llm::runtime::EagleBuffers::Inputs"><span class="n"><span class="pre">Inputs</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">draftBuffers</span></span>, <a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">EagleModule</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">eagleModule</span></span>, <a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManagerE" title="tensorrt_llm::runtime::BufferManager"><span class="n"><span class="pre">BufferManager</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">manager</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4I0ENK12tensorrt_llm7runtime12EagleBuffers13setFromInputsEvRK13RequestVectorRK13RequestVector10SizeType32RK7ITensorRKN12EagleBuffers6InputsERKN7runtime11EagleModuleERKN7runtime13BufferManagerE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-members">Private Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers20scanTempStorageBytesE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers20scanTempStorageBytesE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers20scanTempStorageBytesE"></span><span id="tensorrt_llm::runtime::EagleBuffers::scanTempStorageBytes__std::s"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1a7a3ee7c9131da3f1b780800ce36efb94"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">scanTempStorageBytes</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers20scanTempStorageBytesE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers22reduceTempStorageBytesE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers22reduceTempStorageBytesE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers22reduceTempStorageBytesE"></span><span id="tensorrt_llm::runtime::EagleBuffers::reduceTempStorageBytes__std::s"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1acf18f1113acf2467aba8c067f70f1b6d"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">reduceTempStorageBytes</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers22reduceTempStorageBytesE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers26mDefaultPosteriorThresholdE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers26mDefaultPosteriorThresholdE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers26mDefaultPosteriorThresholdE"></span><span id="tensorrt_llm::runtime::EagleBuffers::mDefaultPosteriorThreshold__float"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1a3e0a04cc62f07fb2a32d6db1a5fe2497"></span><span class="kt"><span class="pre">float</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mDefaultPosteriorThreshold</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0.09f</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers26mDefaultPosteriorThresholdE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers17mDoGreedySamplingE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers17mDoGreedySamplingE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers17mDoGreedySamplingE"></span><span id="tensorrt_llm::runtime::EagleBuffers::mDoGreedySampling__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1ad92b5ea646e7c2781f74cd6cfa245ccc"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mDoGreedySampling</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="k"><span class="pre">true</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers17mDoGreedySamplingE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers21scanReduceTempStorageE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers21scanReduceTempStorageE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers21scanReduceTempStorageE"></span><span id="tensorrt_llm::runtime::EagleBuffers::scanReduceTempStorage__BufferPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1ae0fefea43ce6381642eb2d7180aee23b"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9BufferPtrE" title="tensorrt_llm::runtime::EagleBuffers::BufferPtr"><span class="n"><span class="pre">BufferPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">scanReduceTempStorage</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers21scanReduceTempStorageE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers23cumSumGenerationLengthsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers23cumSumGenerationLengthsE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers23cumSumGenerationLengthsE"></span><span id="tensorrt_llm::runtime::EagleBuffers::cumSumGenerationLengths__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1a6321256e7e048b36f6dce688a6bc1bf0"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">cumSumGenerationLengths</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers23cumSumGenerationLengthsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers19maxGenerationLengthE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers19maxGenerationLengthE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers19maxGenerationLengthE"></span><span id="tensorrt_llm::runtime::EagleBuffers::maxGenerationLength__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1adc072a753d723b4359064be3c9382c1e"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">maxGenerationLength</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers19maxGenerationLengthE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers28chunkedContextNextTokensHostE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers28chunkedContextNextTokensHostE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers28chunkedContextNextTokensHostE"></span><span id="tensorrt_llm::runtime::EagleBuffers::chunkedContextNextTokensHost__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1aae141d4bb117b247d0bea7252a851fea"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">chunkedContextNextTokensHost</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers28chunkedContextNextTokensHostE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers18greedySamplingHostE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers18greedySamplingHostE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers18greedySamplingHostE"></span><span id="tensorrt_llm::runtime::EagleBuffers::greedySamplingHost__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1ab35b1bfa57573173b16f9ea4924d23fc"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">greedySamplingHost</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers18greedySamplingHostE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers18posteriorAlphaHostE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers18posteriorAlphaHostE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers18posteriorAlphaHostE"></span><span id="tensorrt_llm::runtime::EagleBuffers::posteriorAlphaHost__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1a095132fdf32dd2d709314a0f7db31c3d"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">posteriorAlphaHost</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers18posteriorAlphaHostE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers22posteriorThresholdHostE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers22posteriorThresholdHostE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers22posteriorThresholdHostE"></span><span id="tensorrt_llm::runtime::EagleBuffers::posteriorThresholdHost__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1a1f2a8f1a76200d0397c26328fb445f30"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">posteriorThresholdHost</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers22posteriorThresholdHostE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers13EngineOutputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers13EngineOutputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers13EngineOutputsE"></span><span id="tensorrt_llm::runtime::EagleBuffers::EngineOutputs"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1EngineOutputs"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">EngineOutputs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers13EngineOutputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs15nextDraftTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs15nextDraftTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs15nextDraftTokensE"></span><span id="tensorrt_llm::runtime::EagleBuffers::EngineOutputs::nextDraftTokens__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1EngineOutputs_1ac373f5004578db0db8fa9d94b07fa0ea"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">nextDraftTokens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs15nextDraftTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize, maxDecodingDraftTokens] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs13nextDraftLensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs13nextDraftLensE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs13nextDraftLensE"></span><span id="tensorrt_llm::runtime::EagleBuffers::EngineOutputs::nextDraftLens__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1EngineOutputs_1ab0e5f8a16b4bff93b94aa044e0aa353f"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">nextDraftLens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs13nextDraftLensE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs14nextDraftPathsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs14nextDraftPathsE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs14nextDraftPathsE"></span><span id="tensorrt_llm::runtime::EagleBuffers::EngineOutputs::nextDraftPaths__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1EngineOutputs_1a8b7ccc53fa1f8e13984e2a5e594c7746"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">nextDraftPaths</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs14nextDraftPathsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize, maxNumPaths, maxPathLen] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs14acceptedTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs14acceptedTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs14acceptedTokensE"></span><span id="tensorrt_llm::runtime::EagleBuffers::EngineOutputs::acceptedTokens__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1EngineOutputs_1a776fdd4dc9ac5cc76f35ed41241e2bfb"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">acceptedTokens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs14acceptedTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize, maxPathLen] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs12acceptedLensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs12acceptedLensE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs12acceptedLensE"></span><span id="tensorrt_llm::runtime::EagleBuffers::EngineOutputs::acceptedLens__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1EngineOutputs_1afbdae8feb6eaffee2454743440cc6ab6"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">acceptedLens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs12acceptedLensE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs13acceptedPathsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs13acceptedPathsE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs13acceptedPathsE"></span><span id="tensorrt_llm::runtime::EagleBuffers::EngineOutputs::acceptedPaths__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1EngineOutputs_1ae77cf5bd88086dfa071f16043007d58c"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">acceptedPaths</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs13acceptedPathsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs24chunkedContextNextTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs24chunkedContextNextTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs24chunkedContextNextTokensE"></span><span id="tensorrt_llm::runtime::EagleBuffers::EngineOutputs::chunkedContextNextTokens__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1EngineOutputs_1a73e0fca91f622b5cddfa388b702e5062"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">chunkedContextNextTokens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers13EngineOutputs24chunkedContextNextTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize] </p>
|
||
</dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers6InputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers6InputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers6InputsE"></span><span id="tensorrt_llm::runtime::EagleBuffers::Inputs"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1Inputs"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Inputs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6InputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs6createE10SizeType32RKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers6Inputs6createE10SizeType32RKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers6Inputs6createE10SizeType32RKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE"></span><span id="tensorrt_llm::runtime::EagleBuffers::Inputs::create__SizeType32.runtime::TllmRuntimeCR.runtime::ModelConfigCR.runtime::WorldConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1Inputs_1a77ce3aad763b2d61468658afe2a819a1"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">create</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers10SizeType32E" title="tensorrt_llm::runtime::EagleBuffers::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxNumSequences</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs6createE10SizeType32RKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE" title="tensorrt_llm::runtime::EagleBuffers::Inputs::create::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">TllmRuntime</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">runtime</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs6createE10SizeType32RKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE" title="tensorrt_llm::runtime::EagleBuffers::Inputs::create::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ModelConfig</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs6createE10SizeType32RKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE" title="tensorrt_llm::runtime::EagleBuffers::Inputs::create::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">WorldConfig</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">worldConfig</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs6createE10SizeType32RKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs12temperaturesE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers6Inputs12temperaturesE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers6Inputs12temperaturesE"></span><span id="tensorrt_llm::runtime::EagleBuffers::Inputs::temperatures__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1Inputs_1aaf2e4aef805996ca63f565953b5ec118"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">temperatures</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs12temperaturesE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize] or [numSequences] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs14posteriorAlphaE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers6Inputs14posteriorAlphaE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers6Inputs14posteriorAlphaE"></span><span id="tensorrt_llm::runtime::EagleBuffers::Inputs::posteriorAlpha__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1Inputs_1a582f5fbb06de01beb5ffb16b1494ff34"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">posteriorAlpha</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs14posteriorAlphaE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize] or [numSequences] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs18posteriorThresholdE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers6Inputs18posteriorThresholdE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers6Inputs18posteriorThresholdE"></span><span id="tensorrt_llm::runtime::EagleBuffers::Inputs::posteriorThreshold__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1Inputs_1a28a20ed76c484b1e55ee8721ed434af7"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">posteriorThreshold</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs18posteriorThresholdE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize] or [numSequences] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs16randomDataSampleE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers6Inputs16randomDataSampleE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers6Inputs16randomDataSampleE"></span><span id="tensorrt_llm::runtime::EagleBuffers::Inputs::randomDataSample__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1Inputs_1a502fd59c084cd2116b088575daed8e3f"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">randomDataSample</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs16randomDataSampleE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize] or [numSequences] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs20randomDataValidationE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers6Inputs20randomDataValidationE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers6Inputs20randomDataValidationE"></span><span id="tensorrt_llm::runtime::EagleBuffers::Inputs::randomDataValidation__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1Inputs_1af2b06f77002c7960a14c901c0b5746e1"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">randomDataValidation</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs20randomDataValidationE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize, maxDecodingTokens] or [numSequences, maxDecodingTokens] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs11draftTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers6Inputs11draftTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers6Inputs11draftTokensE"></span><span id="tensorrt_llm::runtime::EagleBuffers::Inputs::draftTokens__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1Inputs_1a6d4cec85660e51cb6edb3842e4bfb9d7"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">draftTokens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs11draftTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize, maxDecodingDraftTokens] or [numSequences, maxDecodingDraftTokens] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs9draftLensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers6Inputs9draftLensE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers6Inputs9draftLensE"></span><span id="tensorrt_llm::runtime::EagleBuffers::Inputs::draftLens__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1Inputs_1ad090ea9e7f1aa7906f125f6c82878c97"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">draftLens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs9draftLensE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize] or [numSequences] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs10draftPathsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers6Inputs10draftPathsE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers6Inputs10draftPathsE"></span><span id="tensorrt_llm::runtime::EagleBuffers::Inputs::draftPaths__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1Inputs_1afcddcc096dd0db8b8170f6759b54d4b8"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">draftPaths</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs10draftPathsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize, maxNumPaths, maxPathLen] or [numSequences, maxNumPaths, maxPathLen] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs29specDecodingGenerationLengthsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers6Inputs29specDecodingGenerationLengthsE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers6Inputs29specDecodingGenerationLengthsE"></span><span id="tensorrt_llm::runtime::EagleBuffers::Inputs::specDecodingGenerationLengths__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1Inputs_1a90989ce7f1f133fe2c2bd90ce5d0ec98"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">specDecodingGenerationLengths</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs29specDecodingGenerationLengthsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize] or [numGenSequences] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs33specDecodingGenerationLengthsHostE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers6Inputs33specDecodingGenerationLengthsHostE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers6Inputs33specDecodingGenerationLengthsHostE"></span><span id="tensorrt_llm::runtime::EagleBuffers::Inputs::specDecodingGenerationLengthsHost__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1Inputs_1a7b5bce27b39c26427043ddda02db0a1e"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">specDecodingGenerationLengthsHost</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs33specDecodingGenerationLengthsHostE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize] or [numGenSequences] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs23specDecodingPackedMasksE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers6Inputs23specDecodingPackedMasksE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers6Inputs23specDecodingPackedMasksE"></span><span id="tensorrt_llm::runtime::EagleBuffers::Inputs::specDecodingPackedMasks__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1Inputs_1ac19500a556ddab1fbbd2c13f3fd7df06"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">specDecodingPackedMasks</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs23specDecodingPackedMasksE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize, maxDecodingTokens, ceil(maxDecodingTokens / 32)] or [numGenSequences, maxDecodingTokens, ceil(maxDecodingTokens / 32)] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs27specDecodingPositionOffsetsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers6Inputs27specDecodingPositionOffsetsE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers6Inputs27specDecodingPositionOffsetsE"></span><span id="tensorrt_llm::runtime::EagleBuffers::Inputs::specDecodingPositionOffsets__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1Inputs_1a40ca6da2217921cca5380be65437c1a0"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">specDecodingPositionOffsets</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs27specDecodingPositionOffsetsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize] or [numGenSequences] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs27eagleNetCtxRequestTypesHostE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers6Inputs27eagleNetCtxRequestTypesHostE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers6Inputs27eagleNetCtxRequestTypesHostE"></span><span id="tensorrt_llm::runtime::EagleBuffers::Inputs::eagleNetCtxRequestTypesHost__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1Inputs_1a32fbad1915e6ef5eb5f96c2e61866f88"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">eagleNetCtxRequestTypesHost</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs27eagleNetCtxRequestTypesHostE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize] or [numSequences] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs29eagleNetCtxContextLengthsHostE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers6Inputs29eagleNetCtxContextLengthsHostE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers6Inputs29eagleNetCtxContextLengthsHostE"></span><span id="tensorrt_llm::runtime::EagleBuffers::Inputs::eagleNetCtxContextLengthsHost__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1Inputs_1ae3e80bfd623785f10f74428d5ba70455"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">eagleNetCtxContextLengthsHost</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs29eagleNetCtxContextLengthsHostE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize] or [numSequences] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs34eagleNetCtxPastKeyValueLengthsHostE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers6Inputs34eagleNetCtxPastKeyValueLengthsHostE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers6Inputs34eagleNetCtxPastKeyValueLengthsHostE"></span><span id="tensorrt_llm::runtime::EagleBuffers::Inputs::eagleNetCtxPastKeyValueLengthsHost__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1Inputs_1a027372839fa228e788837bd68590b9ea"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">eagleNetCtxPastKeyValueLengthsHost</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs34eagleNetCtxPastKeyValueLengthsHostE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize] or [numSequences] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs27eagleNetGenRequestTypesHostE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers6Inputs27eagleNetGenRequestTypesHostE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers6Inputs27eagleNetGenRequestTypesHostE"></span><span id="tensorrt_llm::runtime::EagleBuffers::Inputs::eagleNetGenRequestTypesHost__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1Inputs_1adf92b9f5ff67c1c04fedf0e84e1a961b"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">eagleNetGenRequestTypesHost</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs27eagleNetGenRequestTypesHostE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize] or [numSequences] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs29eagleNetGenContextLengthsHostE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers6Inputs29eagleNetGenContextLengthsHostE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers6Inputs29eagleNetGenContextLengthsHostE"></span><span id="tensorrt_llm::runtime::EagleBuffers::Inputs::eagleNetGenContextLengthsHost__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1Inputs_1a02ae021828f1174e028b64db11da1240"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">eagleNetGenContextLengthsHost</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs29eagleNetGenContextLengthsHostE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize] or [numSequences] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs34eagleNetGenPastKeyValueLengthsHostE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers6Inputs34eagleNetGenPastKeyValueLengthsHostE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers6Inputs34eagleNetGenPastKeyValueLengthsHostE"></span><span id="tensorrt_llm::runtime::EagleBuffers::Inputs::eagleNetGenPastKeyValueLengthsHost__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1Inputs_1abbe9ebb8982d302d6e63a3e651d351ef"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">eagleNetGenPastKeyValueLengthsHost</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs34eagleNetGenPastKeyValueLengthsHostE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize] or [numSequences] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs18inputGenTokensHostE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers6Inputs18inputGenTokensHostE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers6Inputs18inputGenTokensHostE"></span><span id="tensorrt_llm::runtime::EagleBuffers::Inputs::inputGenTokensHost__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1Inputs_1ae7c4598e87c543fdf4cd817cbcaeaae0"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">inputGenTokensHost</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs18inputGenTokensHostE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize * maxDecodingTokens] or [numSequences * maxDecodingTokens] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs24chunkedContextNextTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers6Inputs24chunkedContextNextTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers6Inputs24chunkedContextNextTokensE"></span><span id="tensorrt_llm::runtime::EagleBuffers::Inputs::chunkedContextNextTokens__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1Inputs_1aaee773dde70e6e6631a8b4848e0e5efe"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">chunkedContextNextTokens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs24chunkedContextNextTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize] or [numSequences] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs15useSpecDecodingE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers6Inputs15useSpecDecodingE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers6Inputs15useSpecDecodingE"></span><span id="tensorrt_llm::runtime::EagleBuffers::Inputs::useSpecDecoding__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1Inputs_1ae0fc780a1be8f1844c241ec73f762bcb"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">useSpecDecoding</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs15useSpecDecodingE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[1] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs18useDynamicTreeHostE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime12EagleBuffers6Inputs18useDynamicTreeHostE"></span><span id="_CPPv2N12tensorrt_llm7runtime12EagleBuffers6Inputs18useDynamicTreeHostE"></span><span id="tensorrt_llm::runtime::EagleBuffers::Inputs::useDynamicTreeHost__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1EagleBuffers_1_1Inputs_1acda2031f996b58ffe592de1a2d128e98"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers9TensorPtrE" title="tensorrt_llm::runtime::EagleBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">useDynamicTreeHost</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6Inputs18useDynamicTreeHostE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[1] </p>
|
||
</dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="explicitdrafttokensbuffers-h">
|
||
<h2>explicitDraftTokensBuffers.h<a class="headerlink" href="#explicitdrafttokensbuffers-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffersE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffersE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffersE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">ExplicitDraftTokensBuffers</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffersE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers10SizeType32E">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers10SizeType32E"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers10SizeType32E"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1ae16955c34d34bc7ea599673e766f1575"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">SizeType32</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers10SizeType32E" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers7ITensorE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers7ITensorE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers7ITensorE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1a8bc7539f7377aee16c532859d7555c82"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">ITensor</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers7ITensorE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9BufferPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9BufferPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9BufferPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1a8eaca3cd772329f8f8e89643f031b0dc"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">BufferPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer9SharedPtrE" title="tensorrt_llm::runtime::IBuffer::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9BufferPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1ab9edc81aa29738b47db692fdd0b976ae"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorMapE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorMapE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorMapE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1a707525e8e166f8c9777616891ba80177"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorMap</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime12StringPtrMapE" title="tensorrt_llm::runtime::StringPtrMap"><span class="n"><span class="pre">StringPtrMap</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorMapE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers26ExplicitDraftTokensBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerERKN7runtime11ModelConfigERKN7runtime11WorldConfigERKN8executor14DecodingConfigERKN7runtime11TllmRuntimeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers26ExplicitDraftTokensBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerERKN7runtime11ModelConfigERKN7runtime11WorldConfigERKN8executor14DecodingConfigERKN7runtime11TllmRuntimeE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers26ExplicitDraftTokensBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerERKN7runtime11ModelConfigERKN7runtime11WorldConfigERKN8executor14DecodingConfigERKN7runtime11TllmRuntimeE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::ExplicitDraftTokensBuffers__SizeType32.SizeType32.runtime::BufferManagerCR.runtime::ModelConfigCR.runtime::WorldConfigCR.executor::DecodingConfigCR.runtime::TllmRuntimeCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1aeed1193ceacc9b7813e8f0f087401195"></span><span class="sig-name descname"><span class="n"><span class="pre">ExplicitDraftTokensBuffers</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers10SizeType32E" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxBatchSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers10SizeType32E" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxBeamWidth</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers26ExplicitDraftTokensBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerERKN7runtime11ModelConfigERKN7runtime11WorldConfigERKN8executor14DecodingConfigERKN7runtime11TllmRuntimeE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::ExplicitDraftTokensBuffers::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">BufferManager</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">manager</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers26ExplicitDraftTokensBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerERKN7runtime11ModelConfigERKN7runtime11WorldConfigERKN8executor14DecodingConfigERKN7runtime11TllmRuntimeE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::ExplicitDraftTokensBuffers::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ModelConfig</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers26ExplicitDraftTokensBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerERKN7runtime11ModelConfigERKN7runtime11WorldConfigERKN8executor14DecodingConfigERKN7runtime11TllmRuntimeE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::ExplicitDraftTokensBuffers::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">WorldConfig</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">worldConfig</span></span>, <a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executorE" title="tensorrt_llm::executor"><span class="n"><span class="pre">executor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executor14DecodingConfigE" title="tensorrt_llm::executor::DecodingConfig"><span class="n"><span class="pre">DecodingConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">decodingConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers26ExplicitDraftTokensBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerERKN7runtime11ModelConfigERKN7runtime11WorldConfigERKN8executor14DecodingConfigERKN7runtime11TllmRuntimeE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::ExplicitDraftTokensBuffers::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">TllmRuntime</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">runtime</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers26ExplicitDraftTokensBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerERKN7runtime11ModelConfigERKN7runtime11WorldConfigERKN8executor14DecodingConfigERKN7runtime11TllmRuntimeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers7reshapeE10SizeType3210SizeType32RKN7runtime11ModelConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers7reshapeE10SizeType3210SizeType32RKN7runtime11ModelConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers7reshapeE10SizeType3210SizeType32RKN7runtime11ModelConfigE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::reshape__SizeType32.SizeType32.runtime::ModelConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1a9d1e2809ab99e9bb1868f0bc6945355b"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">reshape</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers10SizeType32E" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">numCtxSequences</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers10SizeType32E" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">numGenSequences</span></span>, <a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE" title="tensorrt_llm::runtime::ModelConfig"><span class="n"><span class="pre">ModelConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelConfig</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers7reshapeE10SizeType3210SizeType32RKN7runtime11ModelConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13setFromInputsE10SizeType3210SizeType32RKN7runtime7ITensorERK7ITensorRKN26ExplicitDraftTokensBuffers6InputsERK7ITensorRKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13setFromInputsE10SizeType3210SizeType32RKN7runtime7ITensorERK7ITensorRKN26ExplicitDraftTokensBuffers6InputsERK7ITensorRKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE"></span><span id="_CPPv2NK12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13setFromInputsE10SizeType3210SizeType32RKN7runtime7ITensorERK7ITensorRKN26ExplicitDraftTokensBuffers6InputsERK7ITensorRKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::setFromInputs__SizeType32.SizeType32.runtime::ITensorCR.ITensorCR.ExplicitDraftTokensBuffers::InputsCR.ITensorCR.runtime::TllmRuntimeCR.runtime::ModelConfigCR.runtime::WorldConfigCRC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1af567a079835d83cc23b010104aa6d68a"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setFromInputs</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers10SizeType32E" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">numCtxSequences</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers10SizeType32E" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">numGenSequences</span></span>, <a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13setFromInputsE10SizeType3210SizeType32RKN7runtime7ITensorERK7ITensorRKN26ExplicitDraftTokensBuffers6InputsERK7ITensorRKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::setFromInputs::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ITensor</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">requestTypes</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers7ITensorE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">seqSlots</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffersE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers"><span class="n"><span class="pre">ExplicitDraftTokensBuffers</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6InputsE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::Inputs"><span class="n"><span class="pre">Inputs</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">decoderBuffers</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers7ITensorE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">contextPositionIds</span></span>, <a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13setFromInputsE10SizeType3210SizeType32RKN7runtime7ITensorERK7ITensorRKN26ExplicitDraftTokensBuffers6InputsERK7ITensorRKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::setFromInputs::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">TllmRuntime</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">runtime</span></span>, <a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13setFromInputsE10SizeType3210SizeType32RKN7runtime7ITensorERK7ITensorRKN26ExplicitDraftTokensBuffers6InputsERK7ITensorRKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::setFromInputs::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ModelConfig</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelConfig</span></span>, <a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13setFromInputsE10SizeType3210SizeType32RKN7runtime7ITensorERK7ITensorRKN26ExplicitDraftTokensBuffers6InputsERK7ITensorRKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::setFromInputs::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">WorldConfig</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">worldConfig</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13setFromInputsE10SizeType3210SizeType32RKN7runtime7ITensorERK7ITensorRKN26ExplicitDraftTokensBuffers6InputsERK7ITensorRKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime26ExplicitDraftTokensBuffers18insertInputTensorsER9TensorMapR9TensorMapRKN7runtime11WorldConfigE">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime26ExplicitDraftTokensBuffers18insertInputTensorsER9TensorMapR9TensorMapRKN7runtime11WorldConfigE"></span><span id="_CPPv2NK12tensorrt_llm7runtime26ExplicitDraftTokensBuffers18insertInputTensorsER9TensorMapR9TensorMapRKN7runtime11WorldConfigE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::insertInputTensors__TensorMapR.TensorMapR.runtime::WorldConfigCRC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1af00050bf6dc0ba3b3051417b8cdc2e0a"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">insertInputTensors</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorMapE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorMap"><span class="n"><span class="pre">TensorMap</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">inputBuffers</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorMapE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorMap"><span class="n"><span class="pre">TensorMap</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">outputBuffers</span></span>, <a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfigE" title="tensorrt_llm::runtime::WorldConfig"><span class="n"><span class="pre">WorldConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">worldConfig</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime26ExplicitDraftTokensBuffers18insertInputTensorsER9TensorMapR9TensorMapRKN7runtime11WorldConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers12engineInputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers12engineInputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers12engineInputsE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::engineInputs__tensorrt_llm::runtime::ExplicitDraftTokensBuffers::EngineInputs"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1ab9694adea98429287d1f04738f715e8b"></span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffersE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers"><span class="n"><span class="pre">ExplicitDraftTokensBuffers</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers12EngineInputsE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::EngineInputs"><span class="n"><span class="pre">EngineInputs</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">engineInputs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers12engineInputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13engineOutputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13engineOutputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13engineOutputsE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::engineOutputs__tensorrt_llm::runtime::ExplicitDraftTokensBuffers::EngineOutputs"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1a89f87ae62483ce734c58a1cc254f7dba"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffersE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers"><span class="n"><span class="pre">ExplicitDraftTokensBuffers</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputsE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::EngineOutputs"><span class="n"><span class="pre">EngineOutputs</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">engineOutputs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13engineOutputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers20scanTempStorageBytesE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers20scanTempStorageBytesE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers20scanTempStorageBytesE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::scanTempStorageBytes__std::s"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1a4221d1ab92b52026d1228577030efb2c"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">scanTempStorageBytes</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers20scanTempStorageBytesE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers15scanTempStorageE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers15scanTempStorageE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers15scanTempStorageE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::scanTempStorage__BufferPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1a8b0159a49cf6861dabc340a37defea9c"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9BufferPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::BufferPtr"><span class="n"><span class="pre">BufferPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">scanTempStorage</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers15scanTempStorageE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers23cumSumGenerationLengthsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers23cumSumGenerationLengthsE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers23cumSumGenerationLengthsE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::cumSumGenerationLengths__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1acdfeec1653fe23e61364380a7fbf7e01"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">cumSumGenerationLengths</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers23cumSumGenerationLengthsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-functions">Private Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0ENK12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13setFromInputsEv10SizeType3210SizeType3210SizeType32RK7ITensorRKN26ExplicitDraftTokensBuffers6InputsERK7ITensorRKN7runtime25ExplicitDraftTokensModuleERKN7runtime10CudaStreamE">
|
||
<span id="_CPPv3I0ENK12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13setFromInputsE10SizeType3210SizeType3210SizeType32RK7ITensorRKN26ExplicitDraftTokensBuffers6InputsERK7ITensorRKN7runtime25ExplicitDraftTokensModuleERKN7runtime10CudaStreamE"></span><span id="_CPPv2I0ENK12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13setFromInputsE10SizeType3210SizeType3210SizeType32RK7ITensorRKN26ExplicitDraftTokensBuffers6InputsERK7ITensorRKN7runtime25ExplicitDraftTokensModuleERKN7runtime10CudaStreamE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1a0d5e9b3f821cac5d410ea942fedf0387"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setFromInputs</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers10SizeType32E" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">numCtxSequences</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers10SizeType32E" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">numGenSequences</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers10SizeType32E" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">vocabSizePadded</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers7ITensorE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">seqSlots</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffersE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers"><span class="n"><span class="pre">ExplicitDraftTokensBuffers</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6InputsE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::Inputs"><span class="n"><span class="pre">Inputs</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">draftBuffers</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers7ITensorE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">contextPositionIds</span></span>, <a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ExplicitDraftTokensModule</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">explicitDraftTokensModule</span></span>, <a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10CudaStreamE" title="tensorrt_llm::runtime::CudaStream"><span class="n"><span class="pre">CudaStream</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">stream</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4I0ENK12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13setFromInputsEv10SizeType3210SizeType3210SizeType32RK7ITensorRKN26ExplicitDraftTokensBuffers6InputsERK7ITensorRKN7runtime25ExplicitDraftTokensModuleERKN7runtime10CudaStreamE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers12EngineInputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers12EngineInputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers12EngineInputsE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::EngineInputs"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1EngineInputs"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">EngineInputs</span></span></span><span class="w"> </span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="k"><span class="pre">public</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffersE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers"><span class="n"><span class="pre">ExplicitDraftTokensBuffers</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6InputsE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::Inputs"><span class="n"><span class="pre">Inputs</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers12EngineInputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers12EngineInputs18requestTypesDeviceE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers12EngineInputs18requestTypesDeviceE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers12EngineInputs18requestTypesDeviceE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::EngineInputs::requestTypesDevice__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1EngineInputs_1a2d8a7bc0941d924fcb3f010cacaccd41"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">requestTypesDevice</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers12EngineInputs18requestTypesDeviceE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[numSequences], on gpu </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers12EngineInputs15positionOffsetsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers12EngineInputs15positionOffsetsE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers12EngineInputs15positionOffsetsE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::EngineInputs::positionOffsets__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1EngineInputs_1a8c983b75148ff16930a4662a41f7f53a"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">positionOffsets</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers12EngineInputs15positionOffsetsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[numGenSequences] </p>
|
||
</dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputsE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::EngineOutputs"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1EngineOutputs"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">EngineOutputs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs21nextGenerationLengthsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs21nextGenerationLengthsE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs21nextGenerationLengthsE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::EngineOutputs::nextGenerationLengths__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1EngineOutputs_1a6b4e603a85134a54aa2b450efef3605c"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">nextGenerationLengths</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs21nextGenerationLengthsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs19nextPositionOffsetsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs19nextPositionOffsetsE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs19nextPositionOffsetsE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::EngineOutputs::nextPositionOffsets__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1EngineOutputs_1a873b44001f468d29c7a54009692b5de4"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">nextPositionOffsets</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs19nextPositionOffsetsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs5masksE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs5masksE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs5masksE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::EngineOutputs::masks__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1EngineOutputs_1affd515e8d2260e3f66e5f7e05868cc7e"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">masks</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs5masksE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize, maxDecodingTokens, maxDecodingTokens], bool </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs15nextDraftTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs15nextDraftTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs15nextDraftTokensE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::EngineOutputs::nextDraftTokens__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1EngineOutputs_1a7847820053d967ae770aa92fda4cd3c4"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">nextDraftTokens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs15nextDraftTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize, maxNumPaths, maxPathLen] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs16nextDraftIndicesE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs16nextDraftIndicesE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs16nextDraftIndicesE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::EngineOutputs::nextDraftIndices__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1EngineOutputs_1a789bcc16137e2159d552c4e01057690e"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">nextDraftIndices</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs16nextDraftIndicesE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize, maxNumPaths, maxPathLen] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs14nextDraftProbsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs14nextDraftProbsE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs14nextDraftProbsE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::EngineOutputs::nextDraftProbs__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1EngineOutputs_1a332e0b1cdf83ce26692e3261e3d789ef"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">nextDraftProbs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs14nextDraftProbsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize, maxNumPaths, maxDraftPathLen, vocabSize] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs14nextFlatTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs14nextFlatTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs14nextFlatTokensE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::EngineOutputs::nextFlatTokens__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1EngineOutputs_1a14791c38f4511bd70a0352cbe1593205"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">nextFlatTokens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs14nextFlatTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize * maxDecodingTokens] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs15bestPathLengthsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs15bestPathLengthsE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs15bestPathLengthsE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::EngineOutputs::bestPathLengths__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1EngineOutputs_1a0228355bd26026e839d1d7ebad0b9850"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">bestPathLengths</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs15bestPathLengthsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs15bestPathIndicesE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs15bestPathIndicesE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs15bestPathIndicesE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::EngineOutputs::bestPathIndices__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1EngineOutputs_1a2495fd1a5ae33ab9c5842393e76df22e"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">bestPathIndices</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs15bestPathIndicesE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs11maxGenTokenE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs11maxGenTokenE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs11maxGenTokenE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::EngineOutputs::maxGenToken__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1EngineOutputs_1a6485ad2ce4769f39046e04c07e6a81d3"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">maxGenToken</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs11maxGenTokenE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[1] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs13totalGenTokenE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs13totalGenTokenE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs13totalGenTokenE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::EngineOutputs::totalGenToken__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1EngineOutputs_1aded5114059314a9d25b91878a33b1a27"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">totalGenToken</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs13totalGenTokenE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[1] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs17packedPositionIdsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs17packedPositionIdsE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs17packedPositionIdsE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::EngineOutputs::packedPositionIds__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1EngineOutputs_1ad47080e6b28ea054c28fcac3a7fb0d7e"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">packedPositionIds</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputs17packedPositionIdsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[batchSize * maxDecodingTokens] </p>
|
||
</dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6InputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6InputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6InputsE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::Inputs"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1Inputs"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Inputs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6InputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Subclassed by <a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1EngineInputs"><span class="std std-ref">tensorrt_llm::runtime::ExplicitDraftTokensBuffers::EngineInputs</span></a></p>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs6createE10SizeType32RKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs6createE10SizeType32RKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs6createE10SizeType32RKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::Inputs::create__SizeType32.runtime::TllmRuntimeCR.runtime::ModelConfigCR.runtime::WorldConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1Inputs_1a7083811f3a43e9f5b98390d2df574480"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">create</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers10SizeType32E" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxNumSequences</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs6createE10SizeType32RKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::Inputs::create::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">TllmRuntime</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">runtime</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs6createE10SizeType32RKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::Inputs::create::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ModelConfig</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs6createE10SizeType32RKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::Inputs::create::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">WorldConfig</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">worldConfig</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs6createE10SizeType32RKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs12temperaturesE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs12temperaturesE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs12temperaturesE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::Inputs::temperatures__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1Inputs_1ae5e91bdb5b475a4b997f442ac337407f"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">temperatures</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs12temperaturesE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs15positionIdsBaseE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs15positionIdsBaseE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs15positionIdsBaseE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::Inputs::positionIdsBase__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1Inputs_1a95c39b0d1bdf3f2dc1e73ec7a4c609dd"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">positionIdsBase</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs15positionIdsBaseE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs17generationLengthsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs17generationLengthsE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs17generationLengthsE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::Inputs::generationLengths__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1Inputs_1ae57a628c81b0b92c11ea515159ebbf77"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">generationLengths</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs17generationLengthsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize] or [numGenSequences] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs16randomDataSampleE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs16randomDataSampleE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs16randomDataSampleE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::Inputs::randomDataSample__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1Inputs_1adbee5ac8d0c326af6066aee363de73cc"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">randomDataSample</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs16randomDataSampleE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs20randomDataValidationE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs20randomDataValidationE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs20randomDataValidationE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::Inputs::randomDataValidation__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1Inputs_1ab82f2a2dd80371b340cfe08c672600bb"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">randomDataValidation</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs20randomDataValidationE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize, maxNumPaths, maxPathDraftLen] or [numGenSequences, maxNumPaths, maxPathDraftLen] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs11draftTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs11draftTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs11draftTokensE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::Inputs::draftTokens__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1Inputs_1a083b6d57fd07a2b358322290753e13c0"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">draftTokens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs11draftTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize, maxNumPaths, maxPathLen] or [numGenSequences, maxNumPaths, maxPathLen] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs12draftIndicesE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs12draftIndicesE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs12draftIndicesE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::Inputs::draftIndices__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1Inputs_1ae73718aeff8a57a933308d684abaa9eb"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">draftIndices</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs12draftIndicesE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize, maxNumPaths, maxPathLen] or [numGenSequences, maxNumPaths, maxPathLen] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs10draftProbsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs10draftProbsE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs10draftProbsE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::Inputs::draftProbs__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1Inputs_1aeabf159b722e68e8171c3dc195a5acca"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">draftProbs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs10draftProbsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize, maxNumPaths, maxPathDraftLen, vocabSize] or [numGenSequences, maxNumPaths, maxPathDraftLen, vocabSize] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs11packedMasksE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs11packedMasksE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs11packedMasksE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::Inputs::packedMasks__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1Inputs_1a58685fde1bf5e57bcf3f1a81fb87a550"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">packedMasks</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs11packedMasksE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize, maxDecodingTokens, ceil(maxDecodingTokens / 32)] or [numGenSequences, maxDecodingTokens, ceil(maxDecodingTokens / 32)] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs11positionIdsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs11positionIdsE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs11positionIdsE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::Inputs::positionIds__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1Inputs_1a89285e502e317080a813cea06dc7cfef"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">positionIds</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs11positionIdsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>[maxBatchSize] or [numGenSequences] </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs16maxGenLengthHostE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs16maxGenLengthHostE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs16maxGenLengthHostE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::Inputs::maxGenLengthHost__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1Inputs_1a895e7fb9779d0440dfce06f2269b01fd"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">maxGenLengthHost</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs16maxGenLengthHostE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs21generationLengthsHostE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs21generationLengthsHostE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs21generationLengthsHostE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::Inputs::generationLengthsHost__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1Inputs_1a04c3c6ce76901450a397d18f0e082bed"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">generationLengthsHost</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs21generationLengthsHostE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs15useSpecDecodingE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs15useSpecDecodingE"></span><span id="_CPPv2N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs15useSpecDecodingE"></span><span id="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::Inputs::useSpecDecoding__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ExplicitDraftTokensBuffers_1_1Inputs_1ad79f539b3d0e47dcdf9f9554a0bcf13c"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers9TensorPtrE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">useSpecDecoding</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6Inputs15useSpecDecodingE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="gptdecoder-h">
|
||
<h2>gptDecoder.h<a class="headerlink" href="#gptdecoder-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm6layersE">
|
||
<span id="_CPPv3N12tensorrt_llm6layersE"></span><span id="_CPPv2N12tensorrt_llm6layersE"></span><span id="tensorrt_llm::layers"></span><span class="target" id="namespacetensorrt__llm_1_1layers"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">layers</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm6layersE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-functions">Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime20getDefaultBatchSlotsEN7runtime10SizeType32E">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime20getDefaultBatchSlotsEN7runtime10SizeType32E"></span><span id="_CPPv2N12tensorrt_llm7runtime20getDefaultBatchSlotsEN7runtime10SizeType32E"></span><span id="tensorrt_llm::runtime::getDefaultBatchSlots__runtime::SizeType32"></span><span class="target" id="gptDecoder_8h_1a4be83ec24d8980ca9d74f63e772669e6"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor14SharedConstPtrE" title="tensorrt_llm::runtime::ITensor::SharedConstPtr"><span class="n"><span class="pre">SharedConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getDefaultBatchSlots</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchSize</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime20getDefaultBatchSlotsEN7runtime10SizeType32E" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Helper function to produce batch slots [0, 1, …, batchSize - 1] for paths that do not explicitly provide batch slots to the decoder. </p>
|
||
</dd></dl>
|
||
|
||
</div>
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime10GptDecoderE">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime10GptDecoderE"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime10GptDecoderE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoder"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">GptDecoder</span></span></span><span class="w"> </span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="k"><span class="pre">public</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11IGptDecoderE" title="tensorrt_llm::runtime::IGptDecoder"><span class="n"><span class="pre">IGptDecoder</span></span></a><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime10GptDecoderE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptDecoder13CudaStreamPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptDecoder13CudaStreamPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptDecoder13CudaStreamPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoder_1a8b5c0f02247cc75b5d121681c1990d3b"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">CudaStreamPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManagerE" title="tensorrt_llm::runtime::BufferManager"><span class="n"><span class="pre">BufferManager</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager13CudaStreamPtrE" title="tensorrt_llm::runtime::BufferManager::CudaStreamPtr"><span class="n"><span class="pre">CudaStreamPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder13CudaStreamPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptDecoder9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptDecoder9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptDecoder9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoder_1ac3e1b92c75c404260a36d3d7e873231b"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder9TensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptDecoder10GptDecoderERKN8executor12DecodingModeE6size_t6size_t6size_t6size_t6size_tRK13CudaStreamPtrNSt10shared_ptrIK25SpeculativeDecodingModuleEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptDecoder10GptDecoderERKN8executor12DecodingModeE6size_t6size_t6size_t6size_t6size_tRK13CudaStreamPtrNSt10shared_ptrIK25SpeculativeDecodingModuleEE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptDecoder10GptDecoderERKN8executor12DecodingModeE6size_t6size_t6size_t6size_t6size_tRK13CudaStreamPtrNSt10shared_ptrIK25SpeculativeDecodingModuleEE"></span><span id="tensorrt_llm::runtime::GptDecoder::GptDecoder__executor::DecodingModeCR.s.s.s.s.s.CudaStreamPtrCR.std::shared_ptr:SpeculativeDecodingModuleC:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoder_1a32447c408fdde4dade4948f894f754cb"></span><span class="sig-name descname"><span class="n"><span class="pre">GptDecoder</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executorE" title="tensorrt_llm::executor"><span class="n"><span class="pre">executor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executor12DecodingModeE" title="tensorrt_llm::executor::DecodingMode"><span class="n"><span class="pre">DecodingMode</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">mode</span></span>, <span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">maxBatchSize</span></span>, <span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">maxBeamWidth</span></span>, <span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">vocabSize</span></span>, <span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">vocabSizePadded</span></span>, <span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">maxSequenceLength</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder13CudaStreamPtrE" title="tensorrt_llm::runtime::GptDecoder::CudaStreamPtr"><span class="n"><span class="pre">CudaStreamPtr</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">stream</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModuleE" title="tensorrt_llm::runtime::SpeculativeDecodingModule"><span class="n"><span class="pre">SpeculativeDecodingModule</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="n sig-param"><span class="pre">speculativeDecodingModule</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">nullptr</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder10GptDecoderERKN8executor12DecodingModeE6size_t6size_t6size_t6size_t6size_tRK13CudaStreamPtrNSt10shared_ptrIK25SpeculativeDecodingModuleEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptDecoder5setupERK14SamplingConfig6size_tRK14TensorConstPtrRKNSt8optionalI14DecodingOutputEERKNSt8optionalIKNSt6vectorIN13decoder_batch7RequestEEEEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptDecoder5setupERK14SamplingConfig6size_tRK14TensorConstPtrRKNSt8optionalI14DecodingOutputEERKNSt8optionalIKNSt6vectorIN13decoder_batch7RequestEEEEE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptDecoder5setupERK14SamplingConfig6size_tRK14TensorConstPtrRKNSt8optionalI14DecodingOutputEERKNSt8optionalIKNSt6vectorIN13decoder_batch7RequestEEEEE"></span><span id="tensorrt_llm::runtime::GptDecoder::setup__SamplingConfigCR.s.TensorConstPtrCR.std::optional:DecodingOutput:CR.std::optional:std::vector:decoder_batch::Request:C:CR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoder_1a5f4053070f354e9a1a4a55b35553a980"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setup</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfigE" title="tensorrt_llm::runtime::SamplingConfig"><span class="n"><span class="pre">SamplingConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">samplingConfig</span></span>, <span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">batchSize</span></span>, <span class="n"><span class="pre">TensorConstPtr</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">batchSlots</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutputE" title="tensorrt_llm::runtime::DecodingOutput"><span class="n"><span class="pre">DecodingOutput</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">output</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">nullopt</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7RequestE" title="tensorrt_llm::runtime::decoder_batch::Request"><span class="n"><span class="pre">Request</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">requests</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">nullopt</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder5setupERK14SamplingConfig6size_tRK14TensorConstPtrRKNSt8optionalI14DecodingOutputEERKNSt8optionalIKNSt6vectorIN13decoder_batch7RequestEEEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptDecoder12forwardAsyncER14DecodingOutputRK13DecodingInput">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptDecoder12forwardAsyncER14DecodingOutputRK13DecodingInput"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptDecoder12forwardAsyncER14DecodingOutputRK13DecodingInput"></span><span id="tensorrt_llm::runtime::GptDecoder::forwardAsync__DecodingOutputR.DecodingInputCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoder_1af707cd75a441ad7c639536c75459c65e"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">forwardAsync</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutputE" title="tensorrt_llm::runtime::DecodingOutput"><span class="n"><span class="pre">DecodingOutput</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">output</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInputE" title="tensorrt_llm::runtime::DecodingInput"><span class="n"><span class="pre">DecodingInput</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">input</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder12forwardAsyncER14DecodingOutputRK13DecodingInput" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptDecoder11forwardSyncER14DecodingOutputRK13DecodingInput">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptDecoder11forwardSyncER14DecodingOutputRK13DecodingInput"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptDecoder11forwardSyncER14DecodingOutputRK13DecodingInput"></span><span id="tensorrt_llm::runtime::GptDecoder::forwardSync__DecodingOutputR.DecodingInputCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoder_1a5418b9d72fb9dc359ecd5e1963ef6dfa"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">forwardSync</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutputE" title="tensorrt_llm::runtime::DecodingOutput"><span class="n"><span class="pre">DecodingOutput</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">output</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInputE" title="tensorrt_llm::runtime::DecodingInput"><span class="n"><span class="pre">DecodingInput</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">input</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder11forwardSyncER14DecodingOutputRK13DecodingInput" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptDecoder17getSamplingConfigEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptDecoder17getSamplingConfigEv"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptDecoder17getSamplingConfigEv"></span><span id="tensorrt_llm::runtime::GptDecoder::getSamplingConfig"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoder_1af3947dd72df40649ce40d49a5f51c4dd"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfigE" title="tensorrt_llm::runtime::SamplingConfig"><span class="n"><span class="pre">SamplingConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="n"><span class="pre">getSamplingConfig</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder17getSamplingConfigEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptDecoder16disableLookaheadERKNSt8optionalI14SamplingConfigEE10SizeType3214TensorConstPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptDecoder16disableLookaheadERKNSt8optionalI14SamplingConfigEE10SizeType3214TensorConstPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptDecoder16disableLookaheadERKNSt8optionalI14SamplingConfigEE10SizeType3214TensorConstPtr"></span><span id="tensorrt_llm::runtime::GptDecoder::disableLookahead__std::optional:SamplingConfig:CR.SizeType32.TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoder_1aeef976b7acb0e0956b0380f51b8c7044"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">disableLookahead</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfigE" title="tensorrt_llm::runtime::SamplingConfig"><span class="n"><span class="pre">SamplingConfig</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">samplingConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchSize</span></span>, <span class="n"><span class="pre">TensorConstPtr</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">batchSlots</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder16disableLookaheadERKNSt8optionalI14SamplingConfigEE10SizeType3214TensorConstPtr" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-members">Private Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptDecoder8mManagerE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptDecoder8mManagerE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptDecoder8mManagerE"></span><span id="tensorrt_llm::runtime::GptDecoder::mManager__std::shared_ptr:BufferManager:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoder_1a0abebbe60a46d72ef92f97e9b00c56e9"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManagerE" title="tensorrt_llm::runtime::BufferManager"><span class="n"><span class="pre">BufferManager</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mManager</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder8mManagerE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptDecoder19mDynamicDecodeLayerE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptDecoder19mDynamicDecodeLayerE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptDecoder19mDynamicDecodeLayerE"></span><span id="tensorrt_llm::runtime::GptDecoder::mDynamicDecodeLayer__std::shared_ptr:tensorrt_llm::layers::DynamicDecodeLayer:T::"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoder_1ae47acd96842e4ebc4f34af4fc4370746"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm6layersE" title="tensorrt_llm::layers"><span class="n"><span class="pre">layers</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DynamicDecodeLayer</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime10GptDecoderE" title="tensorrt_llm::runtime::GptDecoder::T"><span class="n"><span class="pre">T</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mDynamicDecodeLayer</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder19mDynamicDecodeLayerE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptDecoder23mDecodingLayerWorkspaceE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptDecoder23mDecodingLayerWorkspaceE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptDecoder23mDecodingLayerWorkspaceE"></span><span id="tensorrt_llm::runtime::GptDecoder::mDecodingLayerWorkspace__std::shared_ptr:tensorrt_llm::runtime::DecodingLayerWorkspace:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoder_1a1a7bb40360534e2fa537ee4ca7122d39"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DecodingLayerWorkspace</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mDecodingLayerWorkspace</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder23mDecodingLayerWorkspaceE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptDecoder15mSamplingConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptDecoder15mSamplingConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptDecoder15mSamplingConfigE"></span><span id="tensorrt_llm::runtime::GptDecoder::mSamplingConfig__SamplingConfig"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoder_1ac921712531dd8bb357fe6787f036db97"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfigE" title="tensorrt_llm::runtime::SamplingConfig"><span class="n"><span class="pre">SamplingConfig</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mSamplingConfig</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder15mSamplingConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptDecoder13mMaxBatchSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptDecoder13mMaxBatchSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptDecoder13mMaxBatchSizeE"></span><span id="tensorrt_llm::runtime::GptDecoder::mMaxBatchSize__s"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoder_1a44386435d3e8b7eabd481c1df21ae61f"></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mMaxBatchSize</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder13mMaxBatchSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptDecoder10mVocabSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptDecoder10mVocabSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptDecoder10mVocabSizeE"></span><span id="tensorrt_llm::runtime::GptDecoder::mVocabSize__s"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoder_1a0d89d86d7ac641b9b1d443e6d22051ce"></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mVocabSize</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder10mVocabSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptDecoder16mVocabSizePaddedE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptDecoder16mVocabSizePaddedE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptDecoder16mVocabSizePaddedE"></span><span id="tensorrt_llm::runtime::GptDecoder::mVocabSizePadded__s"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoder_1a962b3a0289469611233768cf98b2baae"></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mVocabSizePadded</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder16mVocabSizePaddedE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptDecoder13mDecodingModeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptDecoder13mDecodingModeE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptDecoder13mDecodingModeE"></span><span id="tensorrt_llm::runtime::GptDecoder::mDecodingMode__executor::DecodingMode"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoder_1a30eb6c5c819fe128207960862c893d02"></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executorE" title="tensorrt_llm::executor"><span class="n"><span class="pre">executor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executor12DecodingModeE" title="tensorrt_llm::executor::DecodingMode"><span class="n"><span class="pre">DecodingMode</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mDecodingMode</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder13mDecodingModeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11IGptDecoderE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11IGptDecoderE"></span><span id="_CPPv2N12tensorrt_llm7runtime11IGptDecoderE"></span><span id="tensorrt_llm::runtime::IGptDecoder"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoder"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">IGptDecoder</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11IGptDecoderE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Subclassed by <a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1GptDecoder"><span class="std std-ref">tensorrt_llm::runtime::GptDecoder< T ></span></a></p>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11IGptDecoder9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11IGptDecoder9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime11IGptDecoder9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoder_1a2e73d86e70f2cab27226a5339c34b035"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11IGptDecoder9TensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11IGptDecoder14TensorConstPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11IGptDecoder14TensorConstPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime11IGptDecoder14TensorConstPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoder_1a54bda2f28e9738845ad2bc84d4d2335b"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorConstPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor14SharedConstPtrE" title="tensorrt_llm::runtime::ITensor::SharedConstPtr"><span class="n"><span class="pre">SharedConstPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11IGptDecoder14TensorConstPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11IGptDecoderD0Ev">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11IGptDecoderD0Ev"></span><span id="_CPPv2N12tensorrt_llm7runtime11IGptDecoderD0Ev"></span><span id="tensorrt_llm::runtime::IGptDecoder::~IGptDecoder"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoder_1a5ad01b4d1a088b47e56ce6bed969cb57"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">~IGptDecoder</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">default</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11IGptDecoderD0Ev" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11IGptDecoder5setupERK14SamplingConfig6size_tRK14TensorConstPtrRKNSt8optionalI14DecodingOutputEERKNSt8optionalIKNSt6vectorIN13decoder_batch7RequestEEEEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11IGptDecoder5setupERK14SamplingConfig6size_tRK14TensorConstPtrRKNSt8optionalI14DecodingOutputEERKNSt8optionalIKNSt6vectorIN13decoder_batch7RequestEEEEE"></span><span id="_CPPv2N12tensorrt_llm7runtime11IGptDecoder5setupERK14SamplingConfig6size_tRK14TensorConstPtrRKNSt8optionalI14DecodingOutputEERKNSt8optionalIKNSt6vectorIN13decoder_batch7RequestEEEEE"></span><span id="tensorrt_llm::runtime::IGptDecoder::setup__SamplingConfigCR.s.TensorConstPtrCR.std::optional:DecodingOutput:CR.std::optional:std::vector:decoder_batch::Request:C:CR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoder_1af15d3c58f50e58cac3f44d4580e6db84"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setup</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfigE" title="tensorrt_llm::runtime::SamplingConfig"><span class="n"><span class="pre">SamplingConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">samplingConfig</span></span>, <span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">batchSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11IGptDecoder14TensorConstPtrE" title="tensorrt_llm::runtime::IGptDecoder::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">batchSlots</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutputE" title="tensorrt_llm::runtime::DecodingOutput"><span class="n"><span class="pre">DecodingOutput</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">output</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">nullopt</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7RequestE" title="tensorrt_llm::runtime::decoder_batch::Request"><span class="n"><span class="pre">Request</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">requests</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">nullopt</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11IGptDecoder5setupERK14SamplingConfig6size_tRK14TensorConstPtrRKNSt8optionalI14DecodingOutputEERKNSt8optionalIKNSt6vectorIN13decoder_batch7RequestEEEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11IGptDecoder12forwardAsyncER14DecodingOutputRK13DecodingInput">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11IGptDecoder12forwardAsyncER14DecodingOutputRK13DecodingInput"></span><span id="_CPPv2N12tensorrt_llm7runtime11IGptDecoder12forwardAsyncER14DecodingOutputRK13DecodingInput"></span><span id="tensorrt_llm::runtime::IGptDecoder::forwardAsync__DecodingOutputR.DecodingInputCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoder_1a8f87707fea456a6decd013b6b831d336"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">forwardAsync</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutputE" title="tensorrt_llm::runtime::DecodingOutput"><span class="n"><span class="pre">DecodingOutput</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">output</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInputE" title="tensorrt_llm::runtime::DecodingInput"><span class="n"><span class="pre">DecodingInput</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">input</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11IGptDecoder12forwardAsyncER14DecodingOutputRK13DecodingInput" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11IGptDecoder11forwardSyncER14DecodingOutputRK13DecodingInput">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11IGptDecoder11forwardSyncER14DecodingOutputRK13DecodingInput"></span><span id="_CPPv2N12tensorrt_llm7runtime11IGptDecoder11forwardSyncER14DecodingOutputRK13DecodingInput"></span><span id="tensorrt_llm::runtime::IGptDecoder::forwardSync__DecodingOutputR.DecodingInputCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoder_1ae06881ec486e5aadec7d8df477e214e4"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">forwardSync</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutputE" title="tensorrt_llm::runtime::DecodingOutput"><span class="n"><span class="pre">DecodingOutput</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">output</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInputE" title="tensorrt_llm::runtime::DecodingInput"><span class="n"><span class="pre">DecodingInput</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">input</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11IGptDecoder11forwardSyncER14DecodingOutputRK13DecodingInput" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11IGptDecoder17getSamplingConfigEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11IGptDecoder17getSamplingConfigEv"></span><span id="_CPPv2N12tensorrt_llm7runtime11IGptDecoder17getSamplingConfigEv"></span><span id="tensorrt_llm::runtime::IGptDecoder::getSamplingConfig"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoder_1aae2ec40e54ccd288a7c548f09b4f3eb1"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfigE" title="tensorrt_llm::runtime::SamplingConfig"><span class="n"><span class="pre">SamplingConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="n"><span class="pre">getSamplingConfig</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11IGptDecoder17getSamplingConfigEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11IGptDecoder16disableLookaheadERKNSt8optionalI14SamplingConfigEE10SizeType3214TensorConstPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11IGptDecoder16disableLookaheadERKNSt8optionalI14SamplingConfigEE10SizeType3214TensorConstPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime11IGptDecoder16disableLookaheadERKNSt8optionalI14SamplingConfigEE10SizeType3214TensorConstPtr"></span><span id="tensorrt_llm::runtime::IGptDecoder::disableLookahead__std::optional:SamplingConfig:CR.SizeType32.TensorConstPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoder_1a732a15be45afd1f693396e7c88c629af"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">disableLookahead</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfigE" title="tensorrt_llm::runtime::SamplingConfig"><span class="n"><span class="pre">SamplingConfig</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">samplingConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11IGptDecoder14TensorConstPtrE" title="tensorrt_llm::runtime::IGptDecoder::TensorConstPtr"><span class="n"><span class="pre">TensorConstPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchSlots</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11IGptDecoder16disableLookaheadERKNSt8optionalI14SamplingConfigEE10SizeType3214TensorConstPtr" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-functions">Public Static Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11IGptDecoder6createERKN8executor12DecodingModeEN8nvinfer18DataTypeE6size_t6size_t6size_t6size_t6size_tRKN13BufferManager13CudaStreamPtrERKNSt10shared_ptrIK25SpeculativeDecodingModuleEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11IGptDecoder6createERKN8executor12DecodingModeEN8nvinfer18DataTypeE6size_t6size_t6size_t6size_t6size_tRKN13BufferManager13CudaStreamPtrERKNSt10shared_ptrIK25SpeculativeDecodingModuleEE"></span><span id="_CPPv2N12tensorrt_llm7runtime11IGptDecoder6createERKN8executor12DecodingModeEN8nvinfer18DataTypeE6size_t6size_t6size_t6size_t6size_tRKN13BufferManager13CudaStreamPtrERKNSt10shared_ptrIK25SpeculativeDecodingModuleEE"></span><span id="tensorrt_llm::runtime::IGptDecoder::create__executor::DecodingModeCR.nvinfer1::DataType.s.s.s.s.s.BufferManager::CudaStreamPtrCR.std::shared_ptr:SpeculativeDecodingModuleC:CR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoder_1a16c5211cd06c17f8100edacde78c6477"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">unique_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11IGptDecoderE" title="tensorrt_llm::runtime::IGptDecoder"><span class="n"><span class="pre">IGptDecoder</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">create</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executorE" title="tensorrt_llm::executor"><span class="n"><span class="pre">executor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executor12DecodingModeE" title="tensorrt_llm::executor::DecodingMode"><span class="n"><span class="pre">DecodingMode</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">mode</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">dtype</span></span>, <span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">maxBatchSize</span></span>, <span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">maxBeamWidth</span></span>, <span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">vocabSize</span></span>, <span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">vocabSizePadded</span></span>, <span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">maxSequenceLength</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManagerE" title="tensorrt_llm::runtime::BufferManager"><span class="n"><span class="pre">BufferManager</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager13CudaStreamPtrE" title="tensorrt_llm::runtime::BufferManager::CudaStreamPtr"><span class="n"><span class="pre">CudaStreamPtr</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">stream</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModuleE" title="tensorrt_llm::runtime::SpeculativeDecodingModule"><span class="n"><span class="pre">SpeculativeDecodingModule</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">speculativeDecodingModule</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">nullptr</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11IGptDecoder6createERKN8executor12DecodingModeEN8nvinfer18DataTypeE6size_t6size_t6size_t6size_t6size_tRKN13BufferManager13CudaStreamPtrERKNSt10shared_ptrIK25SpeculativeDecodingModuleEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="gptdecoderbatched-h">
|
||
<h2>gptDecoderBatched.h<a class="headerlink" href="#gptdecoderbatched-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1batch__manager"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">batch_manager</span></span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatchedE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatchedE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatchedE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">GptDecoderBatched</span></span></span><span class="w"> </span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="k"><span class="pre">public</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatchedE" title="tensorrt_llm::runtime::IGptDecoderBatched"><span class="n"><span class="pre">IGptDecoderBatched</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatchedE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="docutils container">
|
||
<em>#include <gptDecoderBatched.h></em></div>
|
||
<p>GPT decoder class with support for in-flight batching. </p>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp enum-class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched11ForwardTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched11ForwardTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched11ForwardTypeE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1aefdb7b63a03f888930772f5cf540ac77"></span><span class="k"><span class="pre">enum</span></span><span class="w"> </span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">ForwardType</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched11ForwardTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p><em>Values:</em></p>
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched11ForwardType6kASYNCE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched11ForwardType6kASYNCE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched11ForwardType6kASYNCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1aefdb7b63a03f888930772f5cf540ac77afe4e670be793326f40f9713c57165255"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kASYNC</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched11ForwardType6kASYNCE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched11ForwardType5kSYNCE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched11ForwardType5kSYNCE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched11ForwardType5kSYNCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1aefdb7b63a03f888930772f5cf540ac77ac8a6dec21cd8f4e47c0b68b4b4d81741"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kSYNC</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched11ForwardType5kSYNCE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched13CudaStreamPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched13CudaStreamPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched13CudaStreamPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1aee79b622e8e07d57628f4482d32b340e"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">CudaStreamPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10CudaStreamE" title="tensorrt_llm::runtime::CudaStream"><span class="n"><span class="pre">CudaStream</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched13CudaStreamPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched13LlmRequestPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched13LlmRequestPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched13LlmRequestPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a1050f891cf2cd69288da22e97626d7be"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">LlmRequestPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm13batch_managerE" title="tensorrt_llm::batch_manager"><span class="n"><span class="pre">batch_manager</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">LlmRequest</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched13LlmRequestPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched13RequestVectorE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched13RequestVectorE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched13RequestVectorE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a1a6731b1d1b083cacf268a341ef7d782"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">RequestVector</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched13LlmRequestPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::LlmRequestPtr"><span class="n"><span class="pre">LlmRequestPtr</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched13RequestVectorE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a71918575432e49931d0452cfb4c98a8d"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9TensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14SharedConstPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched14SharedConstPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched14SharedConstPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a2dd6a0a3bdccf9535c9df769033efe2e"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">SharedConstPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor14SharedConstPtrE" title="tensorrt_llm::runtime::ITensor::SharedConstPtr"><span class="n"><span class="pre">SharedConstPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14SharedConstPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched17GptDecoderBatchedENSt6size_tENSt6size_tE13CudaStreamPtrRK23SpeculativeDecodingModeN8nvinfer18DataTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched17GptDecoderBatchedENSt6size_tENSt6size_tE13CudaStreamPtrRK23SpeculativeDecodingModeN8nvinfer18DataTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched17GptDecoderBatchedENSt6size_tENSt6size_tE13CudaStreamPtrRK23SpeculativeDecodingModeN8nvinfer18DataTypeE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::GptDecoderBatched__std::s.std::s.CudaStreamPtr.SpeculativeDecodingModeCR.nvinfer1::DataType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a012b8f52f81587e5c777f86a00b63d02"></span><span class="sig-name descname"><span class="n"><span class="pre">GptDecoderBatched</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">vocabSize</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">vocabSizePadded</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched13CudaStreamPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::CudaStreamPtr"><span class="n"><span class="pre">CudaStreamPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">stream</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingModeE" title="tensorrt_llm::runtime::SpeculativeDecodingMode"><span class="n"><span class="pre">SpeculativeDecodingMode</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">speculativeDecodingMode</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">dtype</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched17GptDecoderBatchedENSt6size_tENSt6size_tE13CudaStreamPtrRK23SpeculativeDecodingModeN8nvinfer18DataTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched5setupERKN8executor12DecodingModeE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType3210SizeType32N8nvinfer18DataTypeERK11ModelConfig">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched5setupERKN8executor12DecodingModeE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType3210SizeType32N8nvinfer18DataTypeERK11ModelConfig"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched5setupERKN8executor12DecodingModeE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType3210SizeType32N8nvinfer18DataTypeERK11ModelConfig"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::setup__executor::DecodingModeCR.SizeType32.SizeType32.SizeType32.SizeType32.SizeType32.SizeType32.nvinfer1::DataType.ModelConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a60a947773443ffe08532d841e37a7965"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setup</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executorE" title="tensorrt_llm::executor"><span class="n"><span class="pre">executor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executor12DecodingModeE" title="tensorrt_llm::executor::DecodingMode"><span class="n"><span class="pre">DecodingMode</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">mode</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxBatchSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxBeamWidth</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxAttentionWindow</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">sinkTokenLength</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxSequenceLength</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxTokensPerStep</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">dtype</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE" title="tensorrt_llm::runtime::ModelConfig"><span class="n"><span class="pre">ModelConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelConfig</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched5setupERKN8executor12DecodingModeE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType3210SizeType32N8nvinfer18DataTypeERK11ModelConfig" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Setup the decoder before calling <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1a75a4af93b82153c19363ff55ceba5c74"><span class="std std-ref"><span class="pre">forward()</span></span></a></code></p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched24setupExplicitDraftTokensEN26ExplicitDraftTokensBuffers6InputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched24setupExplicitDraftTokensEN26ExplicitDraftTokensBuffers6InputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched24setupExplicitDraftTokensEN26ExplicitDraftTokensBuffers6InputsE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::setupExplicitDraftTokens__ExplicitDraftTokensBuffers::Inputs"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a3ce228b717a80f7637c9c1be965347f6"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setupExplicitDraftTokens</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffersE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers"><span class="n"><span class="pre">ExplicitDraftTokensBuffers</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6InputsE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::Inputs"><span class="n"><span class="pre">Inputs</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">explicitDraftTokensBuffers</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched24setupExplicitDraftTokensEN26ExplicitDraftTokensBuffers6InputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Setup buffers for ExplicitDraftTokens decoding. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched10setupEagleEN12EagleBuffers6InputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched10setupEagleEN12EagleBuffers6InputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched10setupEagleEN12EagleBuffers6InputsE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::setupEagle__EagleBuffers::Inputs"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1aa413023983fdb105d97f20442186c93f"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setupEagle</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffersE" title="tensorrt_llm::runtime::EagleBuffers"><span class="n"><span class="pre">EagleBuffers</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6InputsE" title="tensorrt_llm::runtime::EagleBuffers::Inputs"><span class="n"><span class="pre">Inputs</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">eagleBuffers</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched10setupEagleEN12EagleBuffers6InputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Setup buffers for Eagle decoding. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14setupLookaheadE24LookaheadDecodingBuffers">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched14setupLookaheadE24LookaheadDecodingBuffers"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched14setupLookaheadE24LookaheadDecodingBuffers"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::setupLookahead__LookaheadDecodingBuffers"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a48bf64c4bc4e065ef15b2e69f0486a43"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setupLookahead</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffersE" title="tensorrt_llm::runtime::LookaheadDecodingBuffers"><span class="n"><span class="pre">LookaheadDecodingBuffers</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">lookaheadDecodingBuffers</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14setupLookaheadE24LookaheadDecodingBuffers" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Setup buffers for Lookahead decoding. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched16disableLookaheadE10SizeType32RK13RequestVector">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched16disableLookaheadE10SizeType32RK13RequestVector"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched16disableLookaheadE10SizeType32RK13RequestVector"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::disableLookahead__SizeType32.RequestVectorCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a83b2cea349c843115ae51804f89a6a3e"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">disableLookahead</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxBatchSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched13RequestVectorE" title="tensorrt_llm::runtime::GptDecoderBatched::RequestVector"><span class="n"><span class="pre">RequestVector</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">genRequests</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched16disableLookaheadE10SizeType32RK13RequestVector" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Disable Lookahead decoding. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched8newBatchERK15GenerationInputRK16GenerationOutputRK14SamplingConfigRK11ModelConfig">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched8newBatchERK15GenerationInputRK16GenerationOutputRK14SamplingConfigRK11ModelConfig"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched8newBatchERK15GenerationInputRK16GenerationOutputRK14SamplingConfigRK11ModelConfig"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::newBatch__GenerationInputCR.GenerationOutputCR.SamplingConfigCR.ModelConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1aa2210378b1bf78dd54b01fb81c14f318"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">newBatch</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15GenerationInputE" title="tensorrt_llm::runtime::GenerationInput"><span class="n"><span class="pre">GenerationInput</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">inputs</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16GenerationOutputE" title="tensorrt_llm::runtime::GenerationOutput"><span class="n"><span class="pre">GenerationOutput</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">outputs</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfigE" title="tensorrt_llm::runtime::SamplingConfig"><span class="n"><span class="pre">SamplingConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">samplingConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE" title="tensorrt_llm::runtime::ModelConfig"><span class="n"><span class="pre">ModelConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelConfig</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched8newBatchERK15GenerationInputRK16GenerationOutputRK14SamplingConfigRK11ModelConfig" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Initialize the decoder with new batch of inputs. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched11newRequestsERKNSt6vectorI10SizeType32EERKNSt6vectorIN13decoder_batch7RequestEEERKNSt6vectorI14SamplingConfigEERK11ModelConfig">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched11newRequestsERKNSt6vectorI10SizeType32EERKNSt6vectorIN13decoder_batch7RequestEEERKNSt6vectorI14SamplingConfigEERK11ModelConfig"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched11newRequestsERKNSt6vectorI10SizeType32EERKNSt6vectorIN13decoder_batch7RequestEEERKNSt6vectorI14SamplingConfigEERK11ModelConfig"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::newRequests__std::vector:SizeType32:CR.std::vector:decoder_batch::Request:CR.std::vector:SamplingConfig:CR.ModelConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a95e4ddb02dfe6728744c723724415c50"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">newRequests</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">seqSlots</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7RequestE" title="tensorrt_llm::runtime::decoder_batch::Request"><span class="n"><span class="pre">Request</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">requests</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfigE" title="tensorrt_llm::runtime::SamplingConfig"><span class="n"><span class="pre">SamplingConfig</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">samplingConfigs</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE" title="tensorrt_llm::runtime::ModelConfig"><span class="n"><span class="pre">ModelConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelConfig</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched11newRequestsERKNSt6vectorI10SizeType32EERKNSt6vectorIN13decoder_batch7RequestEEERKNSt6vectorI14SamplingConfigEERK11ModelConfig" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Initialize batched decoder at seqSlots with a new <code class="docutils literal notranslate"><span class="pre">requests</span></code>. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched12forwardAsyncERN13decoder_batch6OutputERKN13decoder_batch5InputE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched12forwardAsyncERN13decoder_batch6OutputERKN13decoder_batch5InputE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched12forwardAsyncERN13decoder_batch6OutputERKN13decoder_batch5InputE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::forwardAsync__decoder_batch::OutputR.decoder_batch::InputCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1ac030b7f9a947273729866c425d7ad459"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="n"><span class="pre">DecoderFinishedEventPtr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">forwardAsync</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch6OutputE" title="tensorrt_llm::runtime::decoder_batch::Output"><span class="n"><span class="pre">Output</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">output</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5InputE" title="tensorrt_llm::runtime::decoder_batch::Input"><span class="n"><span class="pre">Input</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">input</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched12forwardAsyncERN13decoder_batch6OutputERKN13decoder_batch5InputE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Run one step for all requests without blocking the host process and return the token for synchronization. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched11forwardSyncERKN13decoder_batch20DecoderFinishedEventE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched11forwardSyncERKN13decoder_batch20DecoderFinishedEventE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched11forwardSyncERKN13decoder_batch20DecoderFinishedEventE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::forwardSync__decoder_batch::DecoderFinishedEventCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a8269c9be94a7cd3cf790bd8a8ef6f8b6"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">forwardSync</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch20DecoderFinishedEventE" title="tensorrt_llm::runtime::decoder_batch::DecoderFinishedEvent"><span class="n"><span class="pre">DecoderFinishedEvent</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">decoderFinishEvent</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched11forwardSyncERKN13decoder_batch20DecoderFinishedEventE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Wait for the call to <code class="docutils literal notranslate"><span class="pre">forwardAsync</span></code> associated with a token to complete. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched11forwardSyncERKN13decoder_batch20DecoderFinishedEventERN13decoder_batch6OutputERKN13decoder_batch5InputE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched11forwardSyncERKN13decoder_batch20DecoderFinishedEventERN13decoder_batch6OutputERKN13decoder_batch5InputE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched11forwardSyncERKN13decoder_batch20DecoderFinishedEventERN13decoder_batch6OutputERKN13decoder_batch5InputE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::forwardSync__decoder_batch::DecoderFinishedEventCR.decoder_batch::OutputR.decoder_batch::InputCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1add94f299d1ba4bcb07c8ad2d6dd215ff"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">forwardSync</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch20DecoderFinishedEventE" title="tensorrt_llm::runtime::decoder_batch::DecoderFinishedEvent"><span class="n"><span class="pre">DecoderFinishedEvent</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">decoderFinishEvent</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch6OutputE" title="tensorrt_llm::runtime::decoder_batch::Output"><span class="n"><span class="pre">Output</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">output</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5InputE" title="tensorrt_llm::runtime::decoder_batch::Input"><span class="n"><span class="pre">Input</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">input</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched11forwardSyncERKN13decoder_batch20DecoderFinishedEventERN13decoder_batch6OutputERKN13decoder_batch5InputE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Call decoder forwardSync and wait for the call to <code class="docutils literal notranslate"><span class="pre">forwardAsync</span></code> associated with a token to complete. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched12forwardAsyncERN7decoder6OutputERKN7decoder5InputE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched12forwardAsyncERN7decoder6OutputERKN7decoder5InputE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched12forwardAsyncERN7decoder6OutputERKN7decoder5InputE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::forwardAsync__decoder::OutputR.decoder::InputCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a54d3acb68a4d46b91bc0f2878cc0c5b0"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">forwardAsync</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7decoderE" title="tensorrt_llm::runtime::decoder"><span class="n"><span class="pre">decoder</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7decoder6OutputE" title="tensorrt_llm::runtime::decoder::Output"><span class="n"><span class="pre">Output</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">output</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7decoderE" title="tensorrt_llm::runtime::decoder"><span class="n"><span class="pre">decoder</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7decoder5InputE" title="tensorrt_llm::runtime::decoder::Input"><span class="n"><span class="pre">Input</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">input</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched12forwardAsyncERN7decoder6OutputERKN7decoder5InputE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Run one step for all requests without blocking the host thread. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched11forwardSyncEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched11forwardSyncEv"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched11forwardSyncEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::forwardSync"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1aedfdad24e80229922b7d725107210d5b"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">forwardSync</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched11forwardSyncEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Wait for the last call to <code class="docutils literal notranslate"><span class="pre">forwardAsync</span></code> to complete. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched11getFinishedEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime17GptDecoderBatched11getFinishedEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime17GptDecoderBatched11getFinishedEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::getFinishedC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a99c3f1157b389ee0c41fd56cdb56bc6b"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="kt"><span class="pre">bool</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getFinished</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched11getFinishedEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[batchSize], indicators of finished requests </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched16getFinishReasonsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime17GptDecoderBatched16getFinishReasonsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime17GptDecoderBatched16getFinishReasonsEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::getFinishReasonsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a180116f7a61ae199b208d9463950bdfe"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getFinishReasons</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched16getFinishReasonsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[batchSize, beamWidth], FinishedState value, on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched6getIdsE10SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime17GptDecoderBatched6getIdsE10SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime17GptDecoderBatched6getIdsE10SizeType32"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::getIds__SizeType32C"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a8a583a27e09ba63c6afaf957072fd340"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getIds</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchIdx</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched6getIdsE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>batchIdx</strong> – index of the batch </p>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>[maxBeamWidth, maxInputLength + maxNewTokens], contains input token ids and generated token ids without padding for request <code class="docutils literal notranslate"><span class="pre">batchIdx</span></code>, on gpu. In case of beam search, contains the ungathered data. </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched6getIdsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime17GptDecoderBatched6getIdsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime17GptDecoderBatched6getIdsEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::getIdsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1ae2d3d645e2d41b4696e26edd755d934b"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getIds</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched6getIdsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[batchSize, maxBeamWidth, maxInputLength + maxNewTokens], contains input token ids and generated token ids without padding, on gpu. In case of beam search, contains the ungathered data. </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched14getGatheredIdsE10SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime17GptDecoderBatched14getGatheredIdsE10SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime17GptDecoderBatched14getGatheredIdsE10SizeType32"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::getGatheredIds__SizeType32C"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1af0983d15280ce465743a3838d8519af5"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getGatheredIds</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchIdx</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched14getGatheredIdsE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>batchIdx</strong> – index of the batch </p>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>[batchSize, maxBeamWidth, maxInputLength + maxNewTokens], only used for beam search. It contains gathered token ids without padding for request <code class="docutils literal notranslate"><span class="pre">batchIdx</span></code>, on gpu. </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched14getGatheredIdsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime17GptDecoderBatched14getGatheredIdsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime17GptDecoderBatched14getGatheredIdsEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::getGatheredIdsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a94d26c1d1a25882d14728c6ae79081f2"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getGatheredIds</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched14getGatheredIdsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[batchSize, maxBeamWidth, maxInputLength + maxNewTokens], only used for beam search. It contains gathered token ids without padding, on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched8finalizeE10SizeType32RK14SamplingConfigb">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime17GptDecoderBatched8finalizeE10SizeType32RK14SamplingConfigb"></span><span id="_CPPv2NK12tensorrt_llm7runtime17GptDecoderBatched8finalizeE10SizeType32RK14SamplingConfigb"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::finalize__SizeType32.SamplingConfigCR.bC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1ae2579801d6c6b662d91fb30d6075357e"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEventE" title="tensorrt_llm::runtime::CudaEvent"><span class="n"><span class="pre">CudaEvent</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">finalize</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchSlot</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfigE" title="tensorrt_llm::runtime::SamplingConfig"><span class="n"><span class="pre">SamplingConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">samplingConfig</span></span>, <span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">streaming</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched8finalizeE10SizeType32RK14SamplingConfigb" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Gather final beam search results for request <code class="docutils literal notranslate"><span class="pre">batchSlot</span></code>. Result will only be available after event returned. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched8finalizeERK14SamplingConfig">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime17GptDecoderBatched8finalizeERK14SamplingConfig"></span><span id="_CPPv2NK12tensorrt_llm7runtime17GptDecoderBatched8finalizeERK14SamplingConfig"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::finalize__SamplingConfigCRC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1afc7675a23490a92bce3a1288818d524a"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">finalize</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfigE" title="tensorrt_llm::runtime::SamplingConfig"><span class="n"><span class="pre">SamplingConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">samplingConfig</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched8finalizeERK14SamplingConfig" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Gather final beam search results for all requests. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched12getParentIdsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime17GptDecoderBatched12getParentIdsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime17GptDecoderBatched12getParentIdsEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::getParentIdsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a1e737bafbd9b040134c7bd9606c25a9b"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getParentIds</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched12getParentIdsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[batchSize, maxBeamWidth, maxInputLength + maxNewTokens], contains parent ids collected during beam search without padding, on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched14getCumLogProbsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime17GptDecoderBatched14getCumLogProbsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime17GptDecoderBatched14getCumLogProbsEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::getCumLogProbsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1abd4fa1cae920f665314d548c6914b4e3"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getCumLogProbs</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched14getCumLogProbsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[batchSize, maxBeamWidth], cumulative log probabilities (per beam), on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched14getCumLogProbsE10SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime17GptDecoderBatched14getCumLogProbsE10SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime17GptDecoderBatched14getCumLogProbsE10SizeType32"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::getCumLogProbs__SizeType32C"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a4e720f2b6070e3642096e8d8823529d8"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getCumLogProbs</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchIdx</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched14getCumLogProbsE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[maxBeamWidth], cumulative log probabilities (per beam), on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched11getLogProbsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime17GptDecoderBatched11getLogProbsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime17GptDecoderBatched11getLogProbsEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::getLogProbsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a1dd13cbd9a82e1f6a67556bff050c1d5"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getLogProbs</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched11getLogProbsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[batchSize, maxBeamWidth, maxSequenceLength], log probabilities (per beam), on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched11getLogProbsE10SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime17GptDecoderBatched11getLogProbsE10SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime17GptDecoderBatched11getLogProbsE10SizeType32"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::getLogProbs__SizeType32C"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1ae1d95c772b21391f4fd28f2c941203ce"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getLogProbs</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchIdx</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched11getLogProbsE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[maxBeamWidth, maxSequenceLength], log probabilities (per beam), on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched15getAllNewTokensEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime17GptDecoderBatched15getAllNewTokensEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime17GptDecoderBatched15getAllNewTokensEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::getAllNewTokensC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1acd4cdca131e9fbb9236605380c466242"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getAllNewTokens</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched15getAllNewTokensEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Get maxTokensPerStep tokens generated in the last forward pass. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[maxTokensPerStep, batchSize, maxBeamWidth], tokens generated in last forward pass, on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched12getNewTokensE10SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime17GptDecoderBatched12getNewTokensE10SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime17GptDecoderBatched12getNewTokensE10SizeType32"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::getNewTokens__SizeType32C"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a1b428a12b68ad44049e2ac68dd1a367f"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getNewTokens</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">iter</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched12getNewTokensE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Get tokens generated in one step of last forward pass. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>iter</strong> – The iteration within [0; maxTokensPerStep) for which to get the tokens </p>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>[batchSize, beamWidth], tokens generated in <code class="docutils literal notranslate"><span class="pre">iter</span></code> (per beam), on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched10getNbStepsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime17GptDecoderBatched10getNbStepsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime17GptDecoderBatched10getNbStepsEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::getNbStepsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a3131818428512dd5e86297dc1e10f2bd"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getNbSteps</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched10getNbStepsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[batchSize], the number of generation steps executed on each request </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched13getNbFinishedEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime17GptDecoderBatched13getNbFinishedEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime17GptDecoderBatched13getNbFinishedEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::getNbFinishedC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a97d1385436fb05b3f1776b6345637698"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getNbFinished</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched13getNbFinishedEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[1], number of finished sequences, in pinned host memory </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched18getNextDraftTokensEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime17GptDecoderBatched18getNextDraftTokensEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime17GptDecoderBatched18getNextDraftTokensEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::getNextDraftTokensC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a8a6a74c9a4618d103abcae296e7ba384"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getNextDraftTokens</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched18getNextDraftTokensEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[batchSize, maxDraftTokens], predicted draft tokens for next step, on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched25getPrevDraftTokensLengthsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime17GptDecoderBatched25getPrevDraftTokensLengthsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime17GptDecoderBatched25getPrevDraftTokensLengthsEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::getPrevDraftTokensLengthsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1abaeb5247260f5b94db786c7e2befc716"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getPrevDraftTokensLengths</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched25getPrevDraftTokensLengthsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[batchSize], predicted draft tokens lengths for previous step, on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched25getNextDraftTokensLengthsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime17GptDecoderBatched25getNextDraftTokensLengthsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime17GptDecoderBatched25getNextDraftTokensLengthsEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::getNextDraftTokensLengthsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a0f955d8d9bce602fa3c4b22b8540aaa5"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getNextDraftTokensLengths</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched25getNextDraftTokensLengthsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[batchSize], predicted draft tokens lengths for next step, on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched24getAcceptedLengthsCumSumEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime17GptDecoderBatched24getAcceptedLengthsCumSumEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime17GptDecoderBatched24getAcceptedLengthsCumSumEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::getAcceptedLengthsCumSumC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a674a3e43d450ed2226443b1ea4112d34"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getAcceptedLengthsCumSum</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched24getAcceptedLengthsCumSumEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[batchSize + 1], exclusive sum of accepted draft token lengths, on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched22getAcceptedPackedPathsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime17GptDecoderBatched22getAcceptedPackedPathsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime17GptDecoderBatched22getAcceptedPackedPathsEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::getAcceptedPackedPathsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1ab3af8a64a6ed40186007211c933bc7d0"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getAcceptedPackedPaths</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched22getAcceptedPackedPathsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[batchSize, maxAcceptedDraftTokensPerStep], accepted paths packed into continuous tensor, on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched15getDecodingModeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime17GptDecoderBatched15getDecodingModeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime17GptDecoderBatched15getDecodingModeEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::getDecodingModeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1afacd9c91ccf04da5d7f13dd007d43ba7"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executorE" title="tensorrt_llm::executor"><span class="n"><span class="pre">executor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executor12DecodingModeE" title="tensorrt_llm::executor::DecodingMode"><span class="n"><span class="pre">DecodingMode</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getDecodingMode</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched15getDecodingModeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-types">Private Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched13GptDecoderPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched13GptDecoderPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched13GptDecoderPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1af8eb97892b63c439e41ccb09adeb2e37"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">GptDecoderPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">unique_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11IGptDecoderE" title="tensorrt_llm::runtime::IGptDecoder"><span class="n"><span class="pre">IGptDecoder</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched13GptDecoderPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched16DecodingInputPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched16DecodingInputPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched16DecodingInputPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1aedfeb64386f5e11a0d4297d3d333ca86"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">DecodingInputPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">unique_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInputE" title="tensorrt_llm::runtime::DecodingInput"><span class="n"><span class="pre">DecodingInput</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched16DecodingInputPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched17DecodingOutputPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched17DecodingOutputPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched17DecodingOutputPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1af058fbb7a903d605c72382a8217fd132"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">DecodingOutputPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">unique_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutputE" title="tensorrt_llm::runtime::DecodingOutput"><span class="n"><span class="pre">DecodingOutput</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched17DecodingOutputPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-functions">Private Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched18postProcessRequestE10SizeType32RK14SamplingConfigb">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime17GptDecoderBatched18postProcessRequestE10SizeType32RK14SamplingConfigb"></span><span id="_CPPv2NK12tensorrt_llm7runtime17GptDecoderBatched18postProcessRequestE10SizeType32RK14SamplingConfigb"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::postProcessRequest__SizeType32.SamplingConfigCR.bC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a97bfb7c9de81b68748d1f4e270da9585"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEventE" title="tensorrt_llm::runtime::CudaEvent"><span class="n"><span class="pre">CudaEvent</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">postProcessRequest</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchIdx</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfigE" title="tensorrt_llm::runtime::SamplingConfig"><span class="n"><span class="pre">SamplingConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">samplingConfig</span></span>, <span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">streaming</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime17GptDecoderBatched18postProcessRequestE10SizeType32RK14SamplingConfigb" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Gather final beam search results for request <code class="docutils literal notranslate"><span class="pre">batchIdx</span></code>. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched10newRequestE10SizeType32RKN13decoder_batch7RequestERK14SamplingConfigRK11ModelConfig">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched10newRequestE10SizeType32RKN13decoder_batch7RequestERK14SamplingConfigRK11ModelConfig"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched10newRequestE10SizeType32RKN13decoder_batch7RequestERK14SamplingConfigRK11ModelConfig"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::newRequest__SizeType32.decoder_batch::RequestCR.SamplingConfigCR.ModelConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a218cd98f480d55a124d752a5482e8622"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">newRequest</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchSlot</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7RequestE" title="tensorrt_llm::runtime::decoder_batch::Request"><span class="n"><span class="pre">Request</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">request</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfigE" title="tensorrt_llm::runtime::SamplingConfig"><span class="n"><span class="pre">SamplingConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">samplingConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE" title="tensorrt_llm::runtime::ModelConfig"><span class="n"><span class="pre">ModelConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelConfig</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched10newRequestE10SizeType32RKN13decoder_batch7RequestERK14SamplingConfigRK11ModelConfig" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Initialize the decoder at <code class="docutils literal notranslate"><span class="pre">batchSlot</span></code> with a new <code class="docutils literal notranslate"><span class="pre">request</span></code>. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched34allocateSpeculativeDecodingBuffersEN8nvinfer18DataTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched34allocateSpeculativeDecodingBuffersEN8nvinfer18DataTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched34allocateSpeculativeDecodingBuffersEN8nvinfer18DataTypeE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::allocateSpeculativeDecodingBuffers__nvinfer1::DataType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1abd0e10c3b574a68a08df50a9e7c5dd03"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">allocateSpeculativeDecodingBuffers</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">dtype</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched34allocateSpeculativeDecodingBuffersEN8nvinfer18DataTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Allocate buffers for speculative decoding. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched24setupSpeculativeDecodingERK11ModelConfig">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched24setupSpeculativeDecodingERK11ModelConfig"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched24setupSpeculativeDecodingERK11ModelConfig"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::setupSpeculativeDecoding__ModelConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a15820bc37b7125f7fcd68bfe06dc86c4"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setupSpeculativeDecoding</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE" title="tensorrt_llm::runtime::ModelConfig"><span class="n"><span class="pre">ModelConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelConfig</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched24setupSpeculativeDecodingERK11ModelConfig" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Setup buffers for speculative decoding. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14setupLookaheadERK11ModelConfig">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched14setupLookaheadERK11ModelConfig"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched14setupLookaheadERK11ModelConfig"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::setupLookahead__ModelConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a698157104704f00e1f999ecb009398e5"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setupLookahead</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE" title="tensorrt_llm::runtime::ModelConfig"><span class="n"><span class="pre">ModelConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelConfig</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14setupLookaheadERK11ModelConfig" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Setup buffers for lookahead decoding. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched29newRequestSpeculativeDecodingE10SizeType32RKN13decoder_batch7RequestERK14SamplingConfigRK11ModelConfig">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched29newRequestSpeculativeDecodingE10SizeType32RKN13decoder_batch7RequestERK14SamplingConfigRK11ModelConfig"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched29newRequestSpeculativeDecodingE10SizeType32RKN13decoder_batch7RequestERK14SamplingConfigRK11ModelConfig"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::newRequestSpeculativeDecoding__SizeType32.decoder_batch::RequestCR.SamplingConfigCR.ModelConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a7f33b5eca1adf0c7dcf4ac583d1e1c40"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">newRequestSpeculativeDecoding</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchIdx</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7RequestE" title="tensorrt_llm::runtime::decoder_batch::Request"><span class="n"><span class="pre">Request</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">request</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfigE" title="tensorrt_llm::runtime::SamplingConfig"><span class="n"><span class="pre">SamplingConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">samplingConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE" title="tensorrt_llm::runtime::ModelConfig"><span class="n"><span class="pre">ModelConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelConfig</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched29newRequestSpeculativeDecodingE10SizeType32RKN13decoder_batch7RequestERK14SamplingConfigRK11ModelConfig" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Setups decoder internal tensors for new speculative decoding request. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched29newRequestDraftTokensExternalE10SizeType32RKN13decoder_batch7RequestERK14SamplingConfig">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched29newRequestDraftTokensExternalE10SizeType32RKN13decoder_batch7RequestERK14SamplingConfig"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched29newRequestDraftTokensExternalE10SizeType32RKN13decoder_batch7RequestERK14SamplingConfig"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::newRequestDraftTokensExternal__SizeType32.decoder_batch::RequestCR.SamplingConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a29dedc4f5dde1f53c8904ed9cb6c1c02"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">newRequestDraftTokensExternal</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchIdx</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7RequestE" title="tensorrt_llm::runtime::decoder_batch::Request"><span class="n"><span class="pre">Request</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">request</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfigE" title="tensorrt_llm::runtime::SamplingConfig"><span class="n"><span class="pre">SamplingConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">samplingConfig</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched29newRequestDraftTokensExternalE10SizeType32RKN13decoder_batch7RequestERK14SamplingConfig" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Setups decoder internal tensors for new request in Draft model Sps mode. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched16newRequestMedusaE10SizeType32RKN13decoder_batch7RequestE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched16newRequestMedusaE10SizeType32RKN13decoder_batch7RequestE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched16newRequestMedusaE10SizeType32RKN13decoder_batch7RequestE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::newRequestMedusa__SizeType32.decoder_batch::RequestCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a407dcfa8d8f5c2d35edd5a4f443b8ee2"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">newRequestMedusa</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchIdx</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7RequestE" title="tensorrt_llm::runtime::decoder_batch::Request"><span class="n"><span class="pre">Request</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">request</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched16newRequestMedusaE10SizeType32RKN13decoder_batch7RequestE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Setups decoder internal tensors for new Medusa request. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched19newRequestLookaheadE10SizeType32RKN13decoder_batch7RequestE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched19newRequestLookaheadE10SizeType32RKN13decoder_batch7RequestE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched19newRequestLookaheadE10SizeType32RKN13decoder_batch7RequestE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::newRequestLookahead__SizeType32.decoder_batch::RequestCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1aba8623c6421397ef227f9095294f808d"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">newRequestLookahead</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchIdx</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7RequestE" title="tensorrt_llm::runtime::decoder_batch::Request"><span class="n"><span class="pre">Request</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">request</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched19newRequestLookaheadE10SizeType32RKN13decoder_batch7RequestE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Setups decoder internal tensors for new Lookahead request. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched29newRequestExplicitDraftTokensE10SizeType32RKN13decoder_batch7RequestE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched29newRequestExplicitDraftTokensE10SizeType32RKN13decoder_batch7RequestE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched29newRequestExplicitDraftTokensE10SizeType32RKN13decoder_batch7RequestE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::newRequestExplicitDraftTokens__SizeType32.decoder_batch::RequestCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1aee89acc64ed3df7ab0b3d1ba76a3a4d9"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">newRequestExplicitDraftTokens</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchIdx</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7RequestE" title="tensorrt_llm::runtime::decoder_batch::Request"><span class="n"><span class="pre">Request</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">request</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched29newRequestExplicitDraftTokensE10SizeType32RKN13decoder_batch7RequestE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Setups decoder internal tensors for new Explicit draft tokens request. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched15newRequestEagleE10SizeType32RKN13decoder_batch7RequestERK11ModelConfig">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched15newRequestEagleE10SizeType32RKN13decoder_batch7RequestERK11ModelConfig"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched15newRequestEagleE10SizeType32RKN13decoder_batch7RequestERK11ModelConfig"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::newRequestEagle__SizeType32.decoder_batch::RequestCR.ModelConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1ae4d9694a47cb746f44aefb40f526e1c1"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">newRequestEagle</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchIdx</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7RequestE" title="tensorrt_llm::runtime::decoder_batch::Request"><span class="n"><span class="pre">Request</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">request</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE" title="tensorrt_llm::runtime::ModelConfig"><span class="n"><span class="pre">ModelConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelConfig</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched15newRequestEagleE10SizeType32RKN13decoder_batch7RequestERK11ModelConfig" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Setups decoder internal tensors for new Eagle request. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14updateFinishedERKN13decoder_batch20DecoderFinishedEventE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched14updateFinishedERKN13decoder_batch20DecoderFinishedEventE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched14updateFinishedERKN13decoder_batch20DecoderFinishedEventE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::updateFinished__decoder_batch::DecoderFinishedEventCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a96f5ce219c1a91014b93fa457e427987"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">updateFinished</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch20DecoderFinishedEventE" title="tensorrt_llm::runtime::decoder_batch::DecoderFinishedEvent"><span class="n"><span class="pre">DecoderFinishedEvent</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">decoderFinishEvent</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14updateFinishedERKN13decoder_batch20DecoderFinishedEventE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Updates finished state on host for all active requests. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched28setExplicitDraftTokensInputsERKN13decoder_batch5InputE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched28setExplicitDraftTokensInputsERKN13decoder_batch5InputE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched28setExplicitDraftTokensInputsERKN13decoder_batch5InputE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::setExplicitDraftTokensInputs__decoder_batch::InputCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1ab20d2145fd96701343c46c4c7d5177a6"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setExplicitDraftTokensInputs</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5InputE" title="tensorrt_llm::runtime::decoder_batch::Input"><span class="n"><span class="pre">Input</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">input</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched28setExplicitDraftTokensInputsERKN13decoder_batch5InputE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Sets inputs for explicit draft tokens. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14setEagleInputsERKN13decoder_batch5InputE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched14setEagleInputsERKN13decoder_batch5InputE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched14setEagleInputsERKN13decoder_batch5InputE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::setEagleInputs__decoder_batch::InputCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a7e170cbad9f4ca2cd84eb3d828e81142"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setEagleInputs</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5InputE" title="tensorrt_llm::runtime::decoder_batch::Input"><span class="n"><span class="pre">Input</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">input</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14setEagleInputsERKN13decoder_batch5InputE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Sets inputs for eagle decoding. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched15forwardDispatchERN13decoder_batch6OutputERKN13decoder_batch5InputE11ForwardType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched15forwardDispatchERN13decoder_batch6OutputERKN13decoder_batch5InputE11ForwardType"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched15forwardDispatchERN13decoder_batch6OutputERKN13decoder_batch5InputE11ForwardType"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::forwardDispatch__decoder_batch::OutputR.decoder_batch::InputCR.ForwardType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1ad6c3d663d095817ed0f5d537ef420dc0"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">forwardDispatch</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch6OutputE" title="tensorrt_llm::runtime::decoder_batch::Output"><span class="n"><span class="pre">Output</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">output</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5InputE" title="tensorrt_llm::runtime::decoder_batch::Input"><span class="n"><span class="pre">Input</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">input</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched11ForwardTypeE" title="tensorrt_llm::runtime::GptDecoderBatched::ForwardType"><span class="n"><span class="pre">ForwardType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">forwardType</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched15forwardDispatchERN13decoder_batch6OutputERKN13decoder_batch5InputE11ForwardType" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Calls decoders for tokens per engine step. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14forwardDecoderE10SizeType32RN13decoder_batch6OutputERKN13decoder_batch5InputE11ForwardType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched14forwardDecoderE10SizeType32RN13decoder_batch6OutputERKN13decoder_batch5InputE11ForwardType"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched14forwardDecoderE10SizeType32RN13decoder_batch6OutputERKN13decoder_batch5InputE11ForwardType"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::forwardDecoder__SizeType32.decoder_batch::OutputR.decoder_batch::InputCR.ForwardType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a8f5baf19fad027902ea5c287d8f3d44c"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">forwardDecoder</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">step</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch6OutputE" title="tensorrt_llm::runtime::decoder_batch::Output"><span class="n"><span class="pre">Output</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">output</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5InputE" title="tensorrt_llm::runtime::decoder_batch::Input"><span class="n"><span class="pre">Input</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">input</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched11ForwardTypeE" title="tensorrt_llm::runtime::GptDecoderBatched::ForwardType"><span class="n"><span class="pre">ForwardType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">forwardType</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14forwardDecoderE10SizeType32RN13decoder_batch6OutputERKN13decoder_batch5InputE11ForwardType" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Calls decoder for whole batch. </p>
|
||
</dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-members">Private Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched10mVocabSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched10mVocabSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched10mVocabSizeE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mVocabSize__std::sC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a4a5174dcbf8d002f6735acc7e69603bb"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mVocabSize</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched10mVocabSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched16mVocabSizePaddedE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched16mVocabSizePaddedE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched16mVocabSizePaddedE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mVocabSizePadded__std::sC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1aa834161ae7981e650786dc405843d213"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mVocabSizePadded</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched16mVocabSizePaddedE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14mRuntimeStreamE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched14mRuntimeStreamE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched14mRuntimeStreamE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mRuntimeStream__CudaStreamPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a70dfe7d65b7106a7f67350fb46222c32"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched13CudaStreamPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::CudaStreamPtr"><span class="n"><span class="pre">CudaStreamPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mRuntimeStream</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14mRuntimeStreamE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14mDecoderStreamE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched14mDecoderStreamE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched14mDecoderStreamE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mDecoderStream__CudaStreamPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1ab9b46b895f196e01eedbe6d9b3206cd4"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched13CudaStreamPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::CudaStreamPtr"><span class="n"><span class="pre">CudaStreamPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mDecoderStream</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14mDecoderStreamE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14mBufferManagerE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched14mBufferManagerE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched14mBufferManagerE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mBufferManager__BufferManager"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a6d0d3bf1e75a46a0e124ae68caecb588"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManagerE" title="tensorrt_llm::runtime::BufferManager"><span class="n"><span class="pre">BufferManager</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mBufferManager</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14mBufferManagerE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched19mDecoderFinishEventE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched19mDecoderFinishEventE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched19mDecoderFinishEventE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mDecoderFinishEvent__DecoderFinishedEventPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1ac9bfb40dba3f0767a738c1ba01946d6c"></span><span class="n"><span class="pre">DecoderFinishedEventPtr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mDecoderFinishEvent</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched19mDecoderFinishEventE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched13mForwardEventE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched13mForwardEventE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched13mForwardEventE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mForwardEvent__CudaEvent"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a511e4db04a983236daca14726ce7cda3"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEventE" title="tensorrt_llm::runtime::CudaEvent"><span class="n"><span class="pre">CudaEvent</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mForwardEvent</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched13mForwardEventE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched8mDecoderE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched8mDecoderE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched8mDecoderE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mDecoder__GptDecoderPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1acc180102b6c64b88146e253d4070e495"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched13GptDecoderPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::GptDecoderPtr"><span class="n"><span class="pre">GptDecoderPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mDecoder</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched8mDecoderE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched19mJointDecodingInputE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched19mJointDecodingInputE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched19mJointDecodingInputE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mJointDecodingInput__DecodingInputPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a8c5c15ffaec96516c2f2f1be315df2af"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched16DecodingInputPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::DecodingInputPtr"><span class="n"><span class="pre">DecodingInputPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mJointDecodingInput</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched19mJointDecodingInputE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched20mJointDecodingOutputE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched20mJointDecodingOutputE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched20mJointDecodingOutputE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mJointDecodingOutput__DecodingOutputPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a23884dee498498d9d6235636e4fd8844"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched17DecodingOutputPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::DecodingOutputPtr"><span class="n"><span class="pre">DecodingOutputPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mJointDecodingOutput</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched20mJointDecodingOutputE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched8mNbStepsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched8mNbStepsE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched8mNbStepsE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mNbSteps__std::vector:SizeType32:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1ac45746fa366b33e1d178be4392c7bce2"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mNbSteps</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched8mNbStepsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9mFinishedE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched9mFinishedE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched9mFinishedE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mFinished__std::vector:b:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a6dca79637197ea0875f278faedcf0b21"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="kt"><span class="pre">bool</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mFinished</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9mFinishedE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched12mFinishedSumE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched12mFinishedSumE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched12mFinishedSumE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mFinishedSum__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a0092b62036d968c257d2e2746fabc109"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mFinishedSum</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched12mFinishedSumE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched13mMaxNewTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched13mMaxNewTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched13mMaxNewTokensE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mMaxNewTokens__std::vector:SizeType32:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a806b8314f68bf9f187f12934aa3e6c2b"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mMaxNewTokens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched13mMaxNewTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched11mBeamWidthsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched11mBeamWidthsE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched11mBeamWidthsE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mBeamWidths__std::vector:SizeType32:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1ad11f910b144ce07e0dbd5626060d73ed"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mBeamWidths</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched11mBeamWidthsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched24mNumDecodingEngineTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched24mNumDecodingEngineTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched24mNumDecodingEngineTokensE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mNumDecodingEngineTokens__std::vector:SizeType32:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a7e84a695835ce3147a330a186ea2993b"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mNumDecodingEngineTokens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched24mNumDecodingEngineTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14mFinishedStepsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched14mFinishedStepsE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched14mFinishedStepsE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mFinishedSteps__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a242b2e68765226757f19930c4ddc7cb3"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mFinishedSteps</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched14mFinishedStepsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched16mBatchSlotsSetupE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched16mBatchSlotsSetupE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched16mBatchSlotsSetupE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mBatchSlotsSetup__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a7e4b5550b71b11e83e7d5bdb861a13f1"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mBatchSlotsSetup</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched16mBatchSlotsSetupE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched18mBatchSlotsDecoderE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched18mBatchSlotsDecoderE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched18mBatchSlotsDecoderE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mBatchSlotsDecoder__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a6fa2c05824a85eb75e4604ca8b26883a"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mBatchSlotsDecoder</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched18mBatchSlotsDecoderE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched18mMaxSequenceLengthE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched18mMaxSequenceLengthE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched18mMaxSequenceLengthE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mMaxSequenceLength__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1aca1d1bd905b0afc58539673f0849d664"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mMaxSequenceLength</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched18mMaxSequenceLengthE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched19mMaxAttentionWindowE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched19mMaxAttentionWindowE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched19mMaxAttentionWindowE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mMaxAttentionWindow__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a57eaf144a940f7a9a9518a4930211265"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mMaxAttentionWindow</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched19mMaxAttentionWindowE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched16mSinkTokenLengthE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched16mSinkTokenLengthE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched16mSinkTokenLengthE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mSinkTokenLength__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a45d4be33ad305d23fadc84ff1f11b687"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mSinkTokenLength</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched16mSinkTokenLengthE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched16mActualBatchSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched16mActualBatchSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched16mActualBatchSizeE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mActualBatchSize__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a47026a7665237fec464645106c2b21d2"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mActualBatchSize</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched16mActualBatchSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched25mMaxDecodingDecoderTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched25mMaxDecodingDecoderTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched25mMaxDecodingDecoderTokensE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mMaxDecodingDecoderTokens__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a93658a234b881402652d0aa681b6e1e2"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mMaxDecodingDecoderTokens</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched25mMaxDecodingDecoderTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched24mMaxDecodingEngineTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched24mMaxDecodingEngineTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched24mMaxDecodingEngineTokensE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mMaxDecodingEngineTokens__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1aa3a9ab7f1bd1d8f7023718da2dc5f0ab"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mMaxDecodingEngineTokens</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched24mMaxDecodingEngineTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched24mSpeculativeDecodingModeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched24mSpeculativeDecodingModeE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched24mSpeculativeDecodingModeE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mSpeculativeDecodingMode__SpeculativeDecodingMode"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a40e56632abb56ce41d63e358a43f2479"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingModeE" title="tensorrt_llm::runtime::SpeculativeDecodingMode"><span class="n"><span class="pre">SpeculativeDecodingMode</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mSpeculativeDecodingMode</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched24mSpeculativeDecodingModeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched13mDecodingModeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched13mDecodingModeE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched13mDecodingModeE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mDecodingMode__executor::DecodingMode"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a0b249459939eaaec00a101875fd4e1ff"></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executorE" title="tensorrt_llm::executor"><span class="n"><span class="pre">executor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executor12DecodingModeE" title="tensorrt_llm::executor::DecodingMode"><span class="n"><span class="pre">DecodingMode</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mDecodingMode</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executorE" title="tensorrt_llm::executor"><span class="n"><span class="pre">executor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executor12DecodingModeE" title="tensorrt_llm::executor::DecodingMode"><span class="n"><span class="pre">DecodingMode</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executor12DecodingMode4AutoEv" title="tensorrt_llm::executor::DecodingMode::Auto"><span class="n"><span class="pre">Auto</span></span></a><span class="p"><span class="pre">(</span></span><span class="p"><span class="pre">)</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched13mDecodingModeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched21mOutputBeamHypothesesE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched21mOutputBeamHypothesesE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched21mOutputBeamHypothesesE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mOutputBeamHypotheses__std::shared_ptr:DecodingOutput::BeamHypotheses:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1ae4b05f96d884b7c8f07d95b9826e3fc7"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutputE" title="tensorrt_llm::runtime::DecodingOutput"><span class="n"><span class="pre">DecodingOutput</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypothesesE" title="tensorrt_llm::runtime::DecodingOutput::BeamHypotheses"><span class="n"><span class="pre">BeamHypotheses</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mOutputBeamHypotheses</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="k"><span class="pre">nullptr</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched21mOutputBeamHypothesesE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched15mCumLogProbsTmpE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched15mCumLogProbsTmpE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched15mCumLogProbsTmpE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mCumLogProbsTmp__DecodingOutput::TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a14da55731f690ad62c8079d616f7b1e7"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutputE" title="tensorrt_llm::runtime::DecodingOutput"><span class="n"><span class="pre">DecodingOutput</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput9TensorPtrE" title="tensorrt_llm::runtime::DecodingOutput::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mCumLogProbsTmp</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched15mCumLogProbsTmpE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched7mNumSMsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime17GptDecoderBatched7mNumSMsE"></span><span id="_CPPv2N12tensorrt_llm7runtime17GptDecoderBatched7mNumSMsE"></span><span id="tensorrt_llm::runtime::GptDecoderBatched::mNumSMs__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatched_1a934cb6c1b54615f99d50ae7295dd8a71"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mNumSMs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime17GptDecoderBatched7mNumSMsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="gptsession-h">
|
||
<h2>gptSession.h<a class="headerlink" href="#gptsession-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1batch__manager"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">batch_manager</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm13batch_manager16kv_cache_managerE">
|
||
<span id="_CPPv3N12tensorrt_llm13batch_manager16kv_cache_managerE"></span><span id="_CPPv2N12tensorrt_llm13batch_manager16kv_cache_managerE"></span><span id="tensorrt_llm::batch_manager::kv_cache_manager"></span><span class="target" id="namespacetensorrt__llm_1_1batch__manager_1_1kv__cache__manager"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kv_cache_manager</span></span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSessionE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSessionE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSessionE"></span><span id="tensorrt_llm::runtime::GptSession"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">GptSession</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSessionE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession9LoggerPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession9LoggerPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession9LoggerPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1abf601de29903bb11af9d3d2976ea08a6"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">LoggerPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ILogger</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession9LoggerPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession10GptSessionERK6ConfigRK11ModelConfigRK11WorldConfigRK9RawEngine9LoggerPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession10GptSessionERK6ConfigRK11ModelConfigRK11WorldConfigRK9RawEngine9LoggerPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession10GptSessionERK6ConfigRK11ModelConfigRK11WorldConfigRK9RawEngine9LoggerPtr"></span><span id="tensorrt_llm::runtime::GptSession::GptSession__ConfigCR.ModelConfigCR.WorldConfigCR.RawEngineCR.LoggerPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a38b95c4ce42c99359b1b0b9ea05bc33a"></span><span class="sig-name descname"><span class="n"><span class="pre">GptSession</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession6ConfigE" title="tensorrt_llm::runtime::GptSession::Config"><span class="n"><span class="pre">Config</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">sessionConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE" title="tensorrt_llm::runtime::ModelConfig"><span class="n"><span class="pre">ModelConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfigE" title="tensorrt_llm::runtime::WorldConfig"><span class="n"><span class="pre">WorldConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">worldConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9RawEngineE" title="tensorrt_llm::runtime::RawEngine"><span class="n"><span class="pre">RawEngine</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">rawEngine</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession9LoggerPtrE" title="tensorrt_llm::runtime::GptSession::LoggerPtr"><span class="n"><span class="pre">LoggerPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">logger</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">nullptr</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession10GptSessionERK6ConfigRK11ModelConfigRK11WorldConfigRK9RawEngine9LoggerPtr" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><ul class="simple">
|
||
<li><p><strong>sessionConfig</strong> – Configuration of the session, </p></li>
|
||
<li><p><strong>modelConfig</strong> – Description of the model, </p></li>
|
||
<li><p><strong>worldConfig</strong> – Description of the environment, </p></li>
|
||
<li><p><strong>rawEngine</strong> – The compiled TensorRT engine, </p></li>
|
||
<li><p><strong>logger</strong> – The optional logger. </p></li>
|
||
</ul>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession10GptSessionERK6ConfigRK11ModelConfigRK11WorldConfigPKvNSt6size_tE9LoggerPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession10GptSessionERK6ConfigRK11ModelConfigRK11WorldConfigPKvNSt6size_tE9LoggerPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession10GptSessionERK6ConfigRK11ModelConfigRK11WorldConfigPKvNSt6size_tE9LoggerPtr"></span><span id="tensorrt_llm::runtime::GptSession::GptSession__ConfigCR.ModelConfigCR.WorldConfigCR.voidCP.std::s.LoggerPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a09aa9cae9db4af5f3d06b92ea03f5116"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">GptSession</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession6ConfigE" title="tensorrt_llm::runtime::GptSession::Config"><span class="n"><span class="pre">Config</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">sessionConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE" title="tensorrt_llm::runtime::ModelConfig"><span class="n"><span class="pre">ModelConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfigE" title="tensorrt_llm::runtime::WorldConfig"><span class="n"><span class="pre">WorldConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">worldConfig</span></span>, <span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="n sig-param"><span class="pre">engineBuffer</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">engineSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession9LoggerPtrE" title="tensorrt_llm::runtime::GptSession::LoggerPtr"><span class="n"><span class="pre">LoggerPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">logger</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">nullptr</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession10GptSessionERK6ConfigRK11ModelConfigRK11WorldConfigPKvNSt6size_tE9LoggerPtr" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession10GptSessionERK6ConfigRK11ModelConfigRK11WorldConfigRKNSt6vectorI7uint8_tEE9LoggerPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession10GptSessionERK6ConfigRK11ModelConfigRK11WorldConfigRKNSt6vectorI7uint8_tEE9LoggerPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession10GptSessionERK6ConfigRK11ModelConfigRK11WorldConfigRKNSt6vectorI7uint8_tEE9LoggerPtr"></span><span id="tensorrt_llm::runtime::GptSession::GptSession__ConfigCR.ModelConfigCR.WorldConfigCR.std::vector:uint8_t:CR.LoggerPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a6b76b6fe682018355d04abfa4c969290"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">GptSession</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession6ConfigE" title="tensorrt_llm::runtime::GptSession::Config"><span class="n"><span class="pre">Config</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">sessionConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE" title="tensorrt_llm::runtime::ModelConfig"><span class="n"><span class="pre">ModelConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfigE" title="tensorrt_llm::runtime::WorldConfig"><span class="n"><span class="pre">WorldConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">worldConfig</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">uint8_t</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">engineBuffer</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession9LoggerPtrE" title="tensorrt_llm::runtime::GptSession::LoggerPtr"><span class="n"><span class="pre">LoggerPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">logger</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">nullptr</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession10GptSessionERK6ConfigRK11ModelConfigRK11WorldConfigRKNSt6vectorI7uint8_tEE9LoggerPtr" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession10GptSessionERK6ConfigRK11ModelConfigRK11WorldConfigRKNSt6stringE9LoggerPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession10GptSessionERK6ConfigRK11ModelConfigRK11WorldConfigRKNSt6stringE9LoggerPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession10GptSessionERK6ConfigRK11ModelConfigRK11WorldConfigRKNSt6stringE9LoggerPtr"></span><span id="tensorrt_llm::runtime::GptSession::GptSession__ConfigCR.ModelConfigCR.WorldConfigCR.ssCR.LoggerPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a41509ce09e144066bf575ef7f800d942"></span><span class="sig-name descname"><span class="n"><span class="pre">GptSession</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession6ConfigE" title="tensorrt_llm::runtime::GptSession::Config"><span class="n"><span class="pre">Config</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">sessionConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE" title="tensorrt_llm::runtime::ModelConfig"><span class="n"><span class="pre">ModelConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfigE" title="tensorrt_llm::runtime::WorldConfig"><span class="n"><span class="pre">WorldConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">worldConfig</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">engineFile</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession9LoggerPtrE" title="tensorrt_llm::runtime::GptSession::LoggerPtr"><span class="n"><span class="pre">LoggerPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">logger</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">nullptr</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession10GptSessionERK6ConfigRK11ModelConfigRK11WorldConfigRKNSt6stringE9LoggerPtr" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10GptSession9getLoggerEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10GptSession9getLoggerEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime10GptSession9getLoggerEv"></span><span id="tensorrt_llm::runtime::GptSession::getLoggerC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a156346e95217701bd7cb2c75f0fe5f67"></span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ILogger</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="n"><span class="pre">getLogger</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession9getLoggerEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10GptSession16getBufferManagerEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10GptSession16getBufferManagerEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime10GptSession16getBufferManagerEv"></span><span id="tensorrt_llm::runtime::GptSession::getBufferManagerC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a281ae2d5c000cc24e9ea4da122687aae"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManagerE" title="tensorrt_llm::runtime::BufferManager"><span class="n"><span class="pre">BufferManager</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="n"><span class="pre">getBufferManager</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession16getBufferManagerEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10GptSession19getRuntimeStreamPtrEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10GptSession19getRuntimeStreamPtrEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime10GptSession19getRuntimeStreamPtrEv"></span><span id="tensorrt_llm::runtime::GptSession::getRuntimeStreamPtrC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a82e840fa867ae5ff24ae2dbe57fb76e8"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManagerE" title="tensorrt_llm::runtime::BufferManager"><span class="n"><span class="pre">BufferManager</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager13CudaStreamPtrE" title="tensorrt_llm::runtime::BufferManager::CudaStreamPtr"><span class="n"><span class="pre">CudaStreamPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getRuntimeStreamPtr</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession19getRuntimeStreamPtrEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10GptSession14getModelConfigEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10GptSession14getModelConfigEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime10GptSession14getModelConfigEv"></span><span id="tensorrt_llm::runtime::GptSession::getModelConfigC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a93d0a8aa55f8d1d6648b2cb21b17c6e9"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE" title="tensorrt_llm::runtime::ModelConfig"><span class="n"><span class="pre">ModelConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="n"><span class="pre">getModelConfig</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession14getModelConfigEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10GptSession14getWorldConfigEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10GptSession14getWorldConfigEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime10GptSession14getWorldConfigEv"></span><span id="tensorrt_llm::runtime::GptSession::getWorldConfigC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1ab0f863be60684fdd658214b3551d587e"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfigE" title="tensorrt_llm::runtime::WorldConfig"><span class="n"><span class="pre">WorldConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="n"><span class="pre">getWorldConfig</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession14getWorldConfigEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10GptSession9getDeviceEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10GptSession9getDeviceEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime10GptSession9getDeviceEv"></span><span id="tensorrt_llm::runtime::GptSession::getDeviceC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a94905fac580c8d0cd5b26b7fe319e853"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">int</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getDevice</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession9getDeviceEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10GptSession20getNormalizeLogProbsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10GptSession20getNormalizeLogProbsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime10GptSession20getNormalizeLogProbsEv"></span><span id="tensorrt_llm::runtime::GptSession::getNormalizeLogProbsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a96662e4f51de07d83847035a7a5a9dc3"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getNormalizeLogProbs</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession20getNormalizeLogProbsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10GptSession18getEngineInspectorEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10GptSession18getEngineInspectorEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime10GptSession18getEngineInspectorEv"></span><span id="tensorrt_llm::runtime::GptSession::getEngineInspectorC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1ac9c2f93d7143e184b4a7c5b17e932f8e"></span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">IEngineInspector</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="n"><span class="pre">getEngineInspector</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession18getEngineInspectorEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10GptSession16getLogitDataTypeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10GptSession16getLogitDataTypeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime10GptSession16getLogitDataTypeEv"></span><span id="tensorrt_llm::runtime::GptSession::getLogitDataTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a591398470162c6e9963d279cb2299a7b"></span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getLogitDataType</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession16getLogitDataTypeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10GptSession17getTensorDataTypeERKNSt6stringE">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10GptSession17getTensorDataTypeERKNSt6stringE"></span><span id="_CPPv2NK12tensorrt_llm7runtime10GptSession17getTensorDataTypeERKNSt6stringE"></span><span id="tensorrt_llm::runtime::GptSession::getTensorDataType__ssCRC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a44e821b47255734ab7a96b6b49360673"></span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getTensorDataType</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">name</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession17getTensorDataTypeERKNSt6stringE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10GptSession14getTensorShapeERKNSt6stringE">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10GptSession14getTensorShapeERKNSt6stringE"></span><span id="_CPPv2NK12tensorrt_llm7runtime10GptSession14getTensorShapeERKNSt6stringE"></span><span id="tensorrt_llm::runtime::GptSession::getTensorShape__ssCRC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a490ae75933eb10803c9d3128779bc63e"></span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">Dims</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getTensorShape</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">name</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession14getTensorShapeERKNSt6stringE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession8generateER16GenerationOutputRK15GenerationInputRK14SamplingConfigKNSt10shared_ptrI18GenerationProfilerEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession8generateER16GenerationOutputRK15GenerationInputRK14SamplingConfigKNSt10shared_ptrI18GenerationProfilerEE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession8generateER16GenerationOutputRK15GenerationInputRK14SamplingConfigKNSt10shared_ptrI18GenerationProfilerEE"></span><span id="tensorrt_llm::runtime::GptSession::generate__GenerationOutputR.GenerationInputCR.SamplingConfigCR.std::shared_ptr:GenerationProfiler:C"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a28a27f9497d62dfba50e9a6c85aa9181"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">generate</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16GenerationOutputE" title="tensorrt_llm::runtime::GenerationOutput"><span class="n"><span class="pre">GenerationOutput</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">outputs</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15GenerationInputE" title="tensorrt_llm::runtime::GenerationInput"><span class="n"><span class="pre">GenerationInput</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">inputs</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfigE" title="tensorrt_llm::runtime::SamplingConfig"><span class="n"><span class="pre">SamplingConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">samplingConfig</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession18GenerationProfilerE" title="tensorrt_llm::runtime::GptSession::GenerationProfiler"><span class="n"><span class="pre">GenerationProfiler</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">generationProfiler</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">nullptr</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession8generateER16GenerationOutputRK15GenerationInputRK14SamplingConfigKNSt10shared_ptrI18GenerationProfilerEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>This function performs the generation loop. </p>
|
||
<p>Given input tensors to read from, output tensors to populate, that member function can be produced or each sequence has reached completion (due to the production will run the generation loop until it reaches the maximum number of tokens that of “end-of-sequence” or a word in the list of “stop words”). The pseudo-code of that function looks like (member function names were changed to keep the presentation simple):</p>
|
||
<div class="highlight-cpp notranslate"><div class="highlight"><pre><span></span><span class="c1">// Have all the sequences in the batch reached completion?</span>
|
||
<span class="kt">bool</span><span class="w"> </span><span class="n">allFinished</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="nb">false</span><span class="p">;</span>
|
||
|
||
<span class="c1">// Until all sequences are finished or the number of steps reaches the limit...</span>
|
||
<span class="k">for</span><span class="w"> </span><span class="p">(</span><span class="kt">int</span><span class="w"> </span><span class="n">step</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="mi">0</span><span class="p">;</span><span class="w"> </span><span class="o">!</span><span class="n">allFinished</span><span class="w"> </span><span class="o">&&</span><span class="w"> </span><span class="n">step</span><span class="w"> </span><span class="o"><</span><span class="w"> </span><span class="n">maxNewTokens</span><span class="p">;</span><span class="w"> </span><span class="o">++</span><span class="n">step</span><span class="p">)</span><span class="w"> </span><span class="p">{</span>
|
||
|
||
<span class="c1">// Trigger the computation of the logits...</span>
|
||
<span class="n">computeLogits</span><span class="p">(...);</span>
|
||
|
||
<span class="c1">// Run the sampling to produce a token (for each active sequence) from the logits.</span>
|
||
<span class="n">allFinished</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="n">generateTokensFromLogits</span><span class="p">(...);</span>
|
||
|
||
<span class="c1">// Callback to stream the output tokens while the generation loop continues.</span>
|
||
<span class="n">onTokenGenerated</span><span class="p">(...);</span>
|
||
<span class="p">}</span>
|
||
</pre></div>
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession16setLayerProfilerEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession16setLayerProfilerEv"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession16setLayerProfilerEv"></span><span id="tensorrt_llm::runtime::GptSession::setLayerProfiler"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a54de04f0902bf58576c3a101e87ec006"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setLayerProfiler</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession16setLayerProfilerEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Set LayerProfiler to collect performance per layer. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10GptSession19getLayerProfileInfoEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10GptSession19getLayerProfileInfoEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime10GptSession19getLayerProfileInfoEv"></span><span id="tensorrt_llm::runtime::GptSession::getLayerProfileInfoC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1abe4713099cf49a1d01f82085afc067ae"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getLayerProfileInfo</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession19getLayerProfileInfoEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Print profile information per layer. </p>
|
||
</dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-types">Private Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession18BaseKVCacheManagerE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession18BaseKVCacheManagerE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession18BaseKVCacheManagerE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1ad7152ca81b534e2a322e9411905473aa"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">BaseKVCacheManager</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm13batch_managerE" title="tensorrt_llm::batch_manager"><span class="n"><span class="pre">batch_manager</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm13batch_manager16kv_cache_managerE" title="tensorrt_llm::batch_manager::kv_cache_manager"><span class="n"><span class="pre">kv_cache_manager</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">BaseKVCacheManager</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession18BaseKVCacheManagerE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession13KvCacheConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession13KvCacheConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession13KvCacheConfigE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a14e12e3a0354f7faa6497a29fb8a4c90"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">KvCacheConfig</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm13batch_managerE" title="tensorrt_llm::batch_manager"><span class="n"><span class="pre">batch_manager</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm13batch_manager16kv_cache_managerE" title="tensorrt_llm::batch_manager::kv_cache_manager"><span class="n"><span class="pre">kv_cache_manager</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">KvCacheConfig</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession13KvCacheConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a3f4d6f00907ec557418c31614d991df4"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession9TensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession22TokenGeneratedCallbackE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession22TokenGeneratedCallbackE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession22TokenGeneratedCallbackE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1ac76592efca13b90289c37eed7e31b3d5"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TokenGeneratedCallback</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">function</span></span><span class="p"><span class="pre"><</span></span><span class="kt"><span class="pre">void</span></span><span class="p"><span class="pre">(</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n"><span class="pre">step</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n"><span class="pre">finished</span></span><span class="p"><span class="pre">)</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession22TokenGeneratedCallbackE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-functions">Private Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession13useCudaGraphsEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession13useCudaGraphsEv"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession13useCudaGraphsEv"></span><span id="tensorrt_llm::runtime::GptSession::useCudaGraphs"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1adae7d88300e62c0e18178dc9962ffd07"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">useCudaGraphs</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession13useCudaGraphsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession15generateBatchedERNSt6vectorI16GenerationOutputEERKNSt6vectorI15GenerationInputEERK14SamplingConfigRK22TokenGeneratedCallbackKNSt10shared_ptrI18GenerationProfilerEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession15generateBatchedERNSt6vectorI16GenerationOutputEERKNSt6vectorI15GenerationInputEERK14SamplingConfigRK22TokenGeneratedCallbackKNSt10shared_ptrI18GenerationProfilerEE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession15generateBatchedERNSt6vectorI16GenerationOutputEERKNSt6vectorI15GenerationInputEERK14SamplingConfigRK22TokenGeneratedCallbackKNSt10shared_ptrI18GenerationProfilerEE"></span><span id="tensorrt_llm::runtime::GptSession::generateBatched__std::vector:GenerationOutput:R.std::vector:GenerationInput:CR.SamplingConfigCR.TokenGeneratedCallbackCR.std::shared_ptr:GenerationProfiler:C"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a8a1954c15cc972107f1dcd5fab0c76eb"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">generateBatched</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16GenerationOutputE" title="tensorrt_llm::runtime::GenerationOutput"><span class="n"><span class="pre">GenerationOutput</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">microBatchesOutputs</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15GenerationInputE" title="tensorrt_llm::runtime::GenerationInput"><span class="n"><span class="pre">GenerationInput</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">microBatchesInputs</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfigE" title="tensorrt_llm::runtime::SamplingConfig"><span class="n"><span class="pre">SamplingConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">samplingConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession22TokenGeneratedCallbackE" title="tensorrt_llm::runtime::GptSession::TokenGeneratedCallback"><span class="n"><span class="pre">TokenGeneratedCallback</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">onTokenGenerated</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession18GenerationProfilerE" title="tensorrt_llm::runtime::GptSession::GenerationProfiler"><span class="n"><span class="pre">GenerationProfiler</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">generationProfiler</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession15generateBatchedERNSt6vectorI16GenerationOutputEERKNSt6vectorI15GenerationInputEERK14SamplingConfigRK22TokenGeneratedCallbackKNSt10shared_ptrI18GenerationProfilerEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession5setupERK6Config">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession5setupERK6Config"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession5setupERK6Config"></span><span id="tensorrt_llm::runtime::GptSession::setup__ConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a1dbeea870aaae467ecfe6961bb2bf259"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setup</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession6ConfigE" title="tensorrt_llm::runtime::GptSession::Config"><span class="n"><span class="pre">Config</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">sessionConfig</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession5setupERK6Config" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession14createContextsEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession14createContextsEv"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession14createContextsEv"></span><span id="tensorrt_llm::runtime::GptSession::createContexts"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1adf43cf5ab563cc203b84f736a5d6b67a"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">createContexts</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession14createContextsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession13createBuffersE10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession13createBuffersE10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession13createBuffersE10SizeType32"></span><span id="tensorrt_llm::runtime::GptSession::createBuffers__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a8c588a6117c9e1e3a178121433831e11"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">createBuffers</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">numMicroBatches</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession13createBuffersE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession14createDecodersE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType32N8nvinfer18DataTypeEb10SizeType32RKN8executor12DecodingModeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession14createDecodersE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType32N8nvinfer18DataTypeEb10SizeType32RKN8executor12DecodingModeE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession14createDecodersE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType32N8nvinfer18DataTypeEb10SizeType32RKN8executor12DecodingModeE"></span><span id="tensorrt_llm::runtime::GptSession::createDecoders__SizeType32.SizeType32.SizeType32.SizeType32.SizeType32.nvinfer1::DataType.b.SizeType32.executor::DecodingModeCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1aa7f263a5ecaffed433342b6fe138db03"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">createDecoders</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">beamWidth</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxAttentionWindow</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">sinkTokenLength</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxSequenceLength</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">logitsType</span></span>, <span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">decoderPerRequest</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">numMicroBatches</span></span>, <a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executorE" title="tensorrt_llm::executor"><span class="n"><span class="pre">executor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executor12DecodingModeE" title="tensorrt_llm::executor::DecodingMode"><span class="n"><span class="pre">DecodingMode</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">decodingMode</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession14createDecodersE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType32N8nvinfer18DataTypeEb10SizeType32RKN8executor12DecodingModeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession20createKvCacheManagerE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType32RK13KvCacheConfig">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession20createKvCacheManagerE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType32RK13KvCacheConfig"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession20createKvCacheManagerE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType32RK13KvCacheConfig"></span><span id="tensorrt_llm::runtime::GptSession::createKvCacheManager__SizeType32.SizeType32.SizeType32.SizeType32.SizeType32.KvCacheConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a27128d2e0548a93948dc9e94eba9af8b"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">createKvCacheManager</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">beamWidth</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxAttentionWindow</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">sinkTokenLength</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxSequenceLength</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession13KvCacheConfigE" title="tensorrt_llm::runtime::GptSession::KvCacheConfig"><span class="n"><span class="pre">KvCacheConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">config</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession20createKvCacheManagerE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType32RK13KvCacheConfig" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession30createCustomAllReduceWorkspaceE10SizeType3210SizeType3210SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession30createCustomAllReduceWorkspaceE10SizeType3210SizeType3210SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession30createCustomAllReduceWorkspaceE10SizeType3210SizeType3210SizeType32"></span><span id="tensorrt_llm::runtime::GptSession::createCustomAllReduceWorkspace__SizeType32.SizeType32.SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a5a52f855f6bcd59677ac5e4f15eb8cae"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">createCustomAllReduceWorkspace</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">beamWidth</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxSequenceLength</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession30createCustomAllReduceWorkspaceE10SizeType3210SizeType3210SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession18executeContextStepERKNSt6vectorI15GenerationInputEERKNSt6vectorI10SizeType32EEPK18BaseKVCacheManager">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession18executeContextStepERKNSt6vectorI15GenerationInputEERKNSt6vectorI10SizeType32EEPK18BaseKVCacheManager"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession18executeContextStepERKNSt6vectorI15GenerationInputEERKNSt6vectorI10SizeType32EEPK18BaseKVCacheManager"></span><span id="tensorrt_llm::runtime::GptSession::executeContextStep__std::vector:GenerationInput:CR.std::vector:SizeType32:CR.BaseKVCacheManagerCP"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a62023e33561d2dbc7899ca0fb668eb71"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">executeContextStep</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15GenerationInputE" title="tensorrt_llm::runtime::GenerationInput"><span class="n"><span class="pre">GenerationInput</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">generationBatchesInputs</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">generationBatchesOffsets</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession18BaseKVCacheManagerE" title="tensorrt_llm::runtime::GptSession::BaseKVCacheManager"><span class="n"><span class="pre">BaseKVCacheManager</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="n sig-param"><span class="pre">kvCacheManager</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession18executeContextStepERKNSt6vectorI15GenerationInputEERKNSt6vectorI10SizeType32EEPK18BaseKVCacheManager" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession21executeGenerationStepE10SizeType32RKNSt6vectorI15GenerationInputEERNSt6vectorI16GenerationOutputEERKNSt6vectorI10SizeType32EEP18BaseKVCacheManagerRNSt6vectorIbEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession21executeGenerationStepE10SizeType32RKNSt6vectorI15GenerationInputEERNSt6vectorI16GenerationOutputEERKNSt6vectorI10SizeType32EEP18BaseKVCacheManagerRNSt6vectorIbEE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession21executeGenerationStepE10SizeType32RKNSt6vectorI15GenerationInputEERNSt6vectorI16GenerationOutputEERKNSt6vectorI10SizeType32EEP18BaseKVCacheManagerRNSt6vectorIbEE"></span><span id="tensorrt_llm::runtime::GptSession::executeGenerationStep__SizeType32.std::vector:GenerationInput:CR.std::vector:GenerationOutput:R.std::vector:SizeType32:CR.BaseKVCacheManagerP.std::vector:b:R"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a455c2a1b4760455d3ad93ebc4d2f04d3"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">executeGenerationStep</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">step</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15GenerationInputE" title="tensorrt_llm::runtime::GenerationInput"><span class="n"><span class="pre">GenerationInput</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">microBatchesInputs</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16GenerationOutputE" title="tensorrt_llm::runtime::GenerationOutput"><span class="n"><span class="pre">GenerationOutput</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">microBatchesOutputs</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">microBatchOffsets</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession18BaseKVCacheManagerE" title="tensorrt_llm::runtime::GptSession::BaseKVCacheManager"><span class="n"><span class="pre">BaseKVCacheManager</span></span></a><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="n sig-param"><span class="pre">kvCacheManager</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="kt"><span class="pre">bool</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">microBatchesFinished</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession21executeGenerationStepE10SizeType32RKNSt6vectorI15GenerationInputEERNSt6vectorI16GenerationOutputEERKNSt6vectorI10SizeType32EEP18BaseKVCacheManagerRNSt6vectorIbEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession16decoderStepAsyncE10SizeType3210SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession16decoderStepAsyncE10SizeType3210SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession16decoderStepAsyncE10SizeType3210SizeType32"></span><span id="tensorrt_llm::runtime::GptSession::decoderStepAsync__SizeType32.SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a1a7c8e400621472d1834f3d1d9a8e4eb"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">decoderStepAsync</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">decoderStep</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">microBatchId</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession16decoderStepAsyncE10SizeType3210SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Execute decoder on last PP rank, receive decoder output on other PP ranks. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession14shouldStopSyncE10SizeType3210SizeType3210SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession14shouldStopSyncE10SizeType3210SizeType3210SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession14shouldStopSyncE10SizeType3210SizeType3210SizeType32"></span><span id="tensorrt_llm::runtime::GptSession::shouldStopSync__SizeType32.SizeType32.SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a2d50e93af6c3a4934742c252c3069866"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">shouldStopSync</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">beamWidth</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">microBatchId</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession14shouldStopSyncE10SizeType3210SizeType3210SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Synchronize with the decoder and return the <code class="docutils literal notranslate"><span class="pre">shouldStop</span></code> flag. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession8finalizeE10SizeType32RK14SamplingConfig">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession8finalizeE10SizeType32RK14SamplingConfig"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession8finalizeE10SizeType32RK14SamplingConfig"></span><span id="tensorrt_llm::runtime::GptSession::finalize__SizeType32.SamplingConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1ae5b39ffba3bad80edf7253b929bfb2b6"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">finalize</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">microBatchId</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfigE" title="tensorrt_llm::runtime::SamplingConfig"><span class="n"><span class="pre">SamplingConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">samplingConfig</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession8finalizeE10SizeType32RK14SamplingConfig" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Collect final output ids and log probs on last PP rank and send them to first PP rank. </p>
|
||
<p>Receives are asynchronous on host, so synchronization is required before access. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession19kvCacheAddSequencesE10SizeType3210SizeType3210SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession19kvCacheAddSequencesE10SizeType3210SizeType3210SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession19kvCacheAddSequencesE10SizeType3210SizeType3210SizeType32"></span><span id="tensorrt_llm::runtime::GptSession::kvCacheAddSequences__SizeType32.SizeType32.SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a1e998d4ede5c2569d102b72f8c0dee54"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kvCacheAddSequences</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">beamWidth</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">microBatchId</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">firstBatchIdx</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession19kvCacheAddSequencesE10SizeType3210SizeType3210SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10GptSession11initDecoderER7ITensorRK15GenerationInputRK16GenerationOutputRK14SamplingConfig10SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10GptSession11initDecoderER7ITensorRK15GenerationInputRK16GenerationOutputRK14SamplingConfig10SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime10GptSession11initDecoderER7ITensorRK15GenerationInputRK16GenerationOutputRK14SamplingConfig10SizeType32"></span><span id="tensorrt_llm::runtime::GptSession::initDecoder__ITensorR.GenerationInputCR.GenerationOutputCR.SamplingConfigCR.SizeType32C"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a53195ff0b61b80f50dc1edaa78bd8dff"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">initDecoder</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">outputIds</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15GenerationInputE" title="tensorrt_llm::runtime::GenerationInput"><span class="n"><span class="pre">GenerationInput</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">inputs</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16GenerationOutputE" title="tensorrt_llm::runtime::GenerationOutput"><span class="n"><span class="pre">GenerationOutput</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">outputs</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfigE" title="tensorrt_llm::runtime::SamplingConfig"><span class="n"><span class="pre">SamplingConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">samplingConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">microBatchId</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession11initDecoderER7ITensorRK15GenerationInputRK16GenerationOutputRK14SamplingConfig10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Populate outputIds and return reference to newTokens tensor. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession30createOnTokenGeneratedCallbackER16GenerationOutput">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession30createOnTokenGeneratedCallbackER16GenerationOutput"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession30createOnTokenGeneratedCallbackER16GenerationOutput"></span><span id="tensorrt_llm::runtime::GptSession::createOnTokenGeneratedCallback__GenerationOutputR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1ac7fab944335f448ad3296ebe39f9ef3f"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession22TokenGeneratedCallbackE" title="tensorrt_llm::runtime::GptSession::TokenGeneratedCallback"><span class="n"><span class="pre">TokenGeneratedCallback</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">createOnTokenGeneratedCallback</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16GenerationOutputE" title="tensorrt_llm::runtime::GenerationOutput"><span class="n"><span class="pre">GenerationOutput</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">outputs</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession30createOnTokenGeneratedCallbackER16GenerationOutput" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10GptSession23shouldUseKVCacheManagerEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10GptSession23shouldUseKVCacheManagerEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime10GptSession23shouldUseKVCacheManagerEv"></span><span id="tensorrt_llm::runtime::GptSession::shouldUseKVCacheManagerC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1aed9e52e3ff437b113592f0bd25e5a43d"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">shouldUseKVCacheManager</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession23shouldUseKVCacheManagerEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-members">Private Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession12mModelConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession12mModelConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession12mModelConfigE"></span><span id="tensorrt_llm::runtime::GptSession::mModelConfig__ModelConfigC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a5afc93709b1d30262510a7399084db52"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE" title="tensorrt_llm::runtime::ModelConfig"><span class="n"><span class="pre">ModelConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mModelConfig</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession12mModelConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession12mWorldConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession12mWorldConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession12mWorldConfigE"></span><span id="tensorrt_llm::runtime::GptSession::mWorldConfig__WorldConfigC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1af712d3cd32f2c0952c6cb5bf1cfece45"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfigE" title="tensorrt_llm::runtime::WorldConfig"><span class="n"><span class="pre">WorldConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mWorldConfig</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession12mWorldConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession7mDeviceE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession7mDeviceE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession7mDeviceE"></span><span id="tensorrt_llm::runtime::GptSession::mDevice__i"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a557d94b8bbfbdc168c417eeae307c25a"></span><span class="kt"><span class="pre">int</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mDevice</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="o"><span class="pre">-</span></span><span class="m"><span class="pre">1</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession7mDeviceE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession13mPipelineCommE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession13mPipelineCommE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession13mPipelineCommE"></span><span id="tensorrt_llm::runtime::GptSession::mPipelineComm__std::shared_ptr:NcclCommunicator:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1afd92278f3aa9e57d02c58f5c5cb1de3a"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">NcclCommunicator</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mPipelineComm</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession13mPipelineCommE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession11mCommStreamE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession11mCommStreamE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession11mCommStreamE"></span><span id="tensorrt_llm::runtime::GptSession::mCommStream__std::shared_ptr:CudaStream:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a14480c0db4c1615ed788836285a3c1e7"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10CudaStreamE" title="tensorrt_llm::runtime::CudaStream"><span class="n"><span class="pre">CudaStream</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mCommStream</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession11mCommStreamE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession10mCommEventE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession10mCommEventE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession10mCommEventE"></span><span id="tensorrt_llm::runtime::GptSession::mCommEvent__CudaEvent"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a0916e45f7ddd9ac0a27eec4897260802"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEventE" title="tensorrt_llm::runtime::CudaEvent"><span class="n"><span class="pre">CudaEvent</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mCommEvent</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession10mCommEventE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession17mAllReduceBuffersE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession17mAllReduceBuffersE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession17mAllReduceBuffersE"></span><span id="tensorrt_llm::runtime::GptSession::mAllReduceBuffers__std::shared_ptr:AllReduceBuffers:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a48d30002b975c6203b8df236a10af76e"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16AllReduceBuffersE" title="tensorrt_llm::runtime::AllReduceBuffers"><span class="n"><span class="pre">AllReduceBuffers</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mAllReduceBuffers</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession17mAllReduceBuffersE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession25mDecoderMaxSequenceLengthE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession25mDecoderMaxSequenceLengthE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession25mDecoderMaxSequenceLengthE"></span><span id="tensorrt_llm::runtime::GptSession::mDecoderMaxSequenceLength__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a25c1a548fe8bc09ebe15201245acd261"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mDecoderMaxSequenceLength</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession25mDecoderMaxSequenceLengthE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession29mDecoderMaxAttentionWindowVecE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession29mDecoderMaxAttentionWindowVecE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession29mDecoderMaxAttentionWindowVecE"></span><span id="tensorrt_llm::runtime::GptSession::mDecoderMaxAttentionWindowVec__std::vector:SizeType32:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a65f9c7f4600b91ee66a44311b73b2328"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mDecoderMaxAttentionWindowVec</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession29mDecoderMaxAttentionWindowVecE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession26mDecoderMaxAttentionWindowE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession26mDecoderMaxAttentionWindowE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession26mDecoderMaxAttentionWindowE"></span><span id="tensorrt_llm::runtime::GptSession::mDecoderMaxAttentionWindow__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a9d39ed9dacff0440edc7b2875eb44ee2"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mDecoderMaxAttentionWindow</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession26mDecoderMaxAttentionWindowE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession23mDecoderSinkTokenLengthE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession23mDecoderSinkTokenLengthE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession23mDecoderSinkTokenLengthE"></span><span id="tensorrt_llm::runtime::GptSession::mDecoderSinkTokenLength__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1adec5713127f2e0c40982690a63d188e5"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mDecoderSinkTokenLength</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession23mDecoderSinkTokenLengthE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession7mLoggerE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession7mLoggerE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession7mLoggerE"></span><span id="tensorrt_llm::runtime::GptSession::mLogger__LoggerPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a4bd2d4d4f201bf17a87ed40c4e29ce55"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession9LoggerPtrE" title="tensorrt_llm::runtime::GptSession::LoggerPtr"><span class="n"><span class="pre">LoggerPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mLogger</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession7mLoggerE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession8mRuntimeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession8mRuntimeE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession8mRuntimeE"></span><span id="tensorrt_llm::runtime::GptSession::mRuntime__std::shared_ptr:TllmRuntime:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1aa97fc0a02f3f28e3e2121e2e9bc10c56"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">TllmRuntime</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mRuntime</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession8mRuntimeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession15mKvCacheManagerE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession15mKvCacheManagerE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession15mKvCacheManagerE"></span><span id="tensorrt_llm::runtime::GptSession::mKvCacheManager__std::shared_ptr:BaseKVCacheManager:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a55a34d445a2f8fa9759555d0a8b59668"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession18BaseKVCacheManagerE" title="tensorrt_llm::runtime::GptSession::BaseKVCacheManager"><span class="n"><span class="pre">BaseKVCacheManager</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mKvCacheManager</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession15mKvCacheManagerE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession17mMicroBatchConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession17mMicroBatchConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession17mMicroBatchConfigE"></span><span id="tensorrt_llm::runtime::GptSession::mMicroBatchConfig__MicroBatchConfig"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a70891459edc318254e61e107ec5edcb2"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession16MicroBatchConfigE" title="tensorrt_llm::runtime::GptSession::MicroBatchConfig"><span class="n"><span class="pre">MicroBatchConfig</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mMicroBatchConfig</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession17mMicroBatchConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession9mDecodersE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession9mDecodersE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession9mDecodersE"></span><span id="tensorrt_llm::runtime::GptSession::mDecoders__std::vector:std::shared_ptr:IStatefulGptDecoder::"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1ab3287006aede8e62f11db44d5f8997e6"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoderE" title="tensorrt_llm::runtime::IStatefulGptDecoder"><span class="n"><span class="pre">IStatefulGptDecoder</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mDecoders</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession9mDecodersE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession8mBuffersE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession8mBuffersE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession8mBuffersE"></span><span id="tensorrt_llm::runtime::GptSession::mBuffers__std::vector:std::shared_ptr:RuntimeBuffers::"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a96b0baf2000c1616dcdcecf836cdd143"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">RuntimeBuffers</span></span><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mBuffers</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession8mBuffersE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession15mReceivedEventsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession15mReceivedEventsE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession15mReceivedEventsE"></span><span id="tensorrt_llm::runtime::GptSession::mReceivedEvents__std::vector:CudaEvent:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a9253a1bccec0b1ab57bab7b7d5b6f5bb"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEventE" title="tensorrt_llm::runtime::CudaEvent"><span class="n"><span class="pre">CudaEvent</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mReceivedEvents</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession15mReceivedEventsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession14mCudaGraphModeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession14mCudaGraphModeE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession14mCudaGraphModeE"></span><span id="tensorrt_llm::runtime::GptSession::mCudaGraphMode__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a1bf104808d61798aeba1c87beeb2c101"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mCudaGraphMode</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="k"><span class="pre">false</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession14mCudaGraphModeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession19mCudaGraphInstancesE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession19mCudaGraphInstancesE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession19mCudaGraphInstancesE"></span><span id="tensorrt_llm::runtime::GptSession::mCudaGraphInstances__std::vector:CudaGraphExecutor:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1abf794ab36d1086fbf43a6a33db6e5181"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession17CudaGraphExecutorE" title="tensorrt_llm::runtime::GptSession::CudaGraphExecutor"><span class="n"><span class="pre">CudaGraphExecutor</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mCudaGraphInstances</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession19mCudaGraphInstancesE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession18mNormalizeLogProbsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession18mNormalizeLogProbsE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession18mNormalizeLogProbsE"></span><span id="tensorrt_llm::runtime::GptSession::mNormalizeLogProbs__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a8fbc240e8add38921dda95a4ab711340"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mNormalizeLogProbs</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">true</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession18mNormalizeLogProbsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-friends">Friends</p>
|
||
<dl>
|
||
<dt class="sig sig-object cpp">
|
||
<em class="property"><span class="pre">friend</span> <span class="pre">class</span></em> <span class="pre">batch_manager::TrtGptModelV1</span></dt>
|
||
</dl>
|
||
|
||
</div>
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession6ConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession6ConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession6ConfigE"></span><span id="tensorrt_llm::runtime::GptSession::Config"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1Config"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Config</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession6ConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="docutils container">
|
||
<em>#include <gptSession.h></em></div>
|
||
<p>Configuration for session execution and buffer sizes. <code class="docutils literal notranslate"><span class="pre">generate</span></code> may be called with batch size and beam width smaller than the configured parameters. </p>
|
||
<p><code class="docutils literal notranslate"><span class="pre">maxBatchSize</span></code> will be divided by the number of micro batches to initialize each batch buffer. </p>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession6Config6ConfigE10SizeType3210SizeType3210SizeType32f">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession6Config6ConfigE10SizeType3210SizeType3210SizeType32f"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession6Config6ConfigE10SizeType3210SizeType3210SizeType32f"></span><span id="tensorrt_llm::runtime::GptSession::Config::Config__SizeType32.SizeType32.SizeType32.float"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1Config_1a21ed90a9aafd8bfe0a027f1758c232e9"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Config</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxBatchSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxBeamWidth</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxSequenceLength</span></span>, <span class="kt"><span class="pre">float</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">gpuWeightsPercent</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">1.0</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession6Config6ConfigE10SizeType3210SizeType3210SizeType32f" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession6Config12maxBatchSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession6Config12maxBatchSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession6Config12maxBatchSizeE"></span><span id="tensorrt_llm::runtime::GptSession::Config::maxBatchSize__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1Config_1a701533e2172a74b123efe37fecb22cec"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">maxBatchSize</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession6Config12maxBatchSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession6Config12maxBeamWidthE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession6Config12maxBeamWidthE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession6Config12maxBeamWidthE"></span><span id="tensorrt_llm::runtime::GptSession::Config::maxBeamWidth__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1Config_1acb85420dfae87a492aee47bcd9630934"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">maxBeamWidth</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession6Config12maxBeamWidthE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession6Config17maxSequenceLengthE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession6Config17maxSequenceLengthE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession6Config17maxSequenceLengthE"></span><span id="tensorrt_llm::runtime::GptSession::Config::maxSequenceLength__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1Config_1aee444c7a95604bc3a8464179600cd9a3"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">maxSequenceLength</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession6Config17maxSequenceLengthE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession6Config17gpuWeightsPercentE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession6Config17gpuWeightsPercentE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession6Config17gpuWeightsPercentE"></span><span id="tensorrt_llm::runtime::GptSession::Config::gpuWeightsPercent__float"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1Config_1a00784de83f9e54fa8e37338dbd68bccf"></span><span class="kt"><span class="pre">float</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">gpuWeightsPercent</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession6Config17gpuWeightsPercentE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession6Config17decoderPerRequestE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession6Config17decoderPerRequestE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession6Config17decoderPerRequestE"></span><span id="tensorrt_llm::runtime::GptSession::Config::decoderPerRequest__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1Config_1a3a5fc9eed846017ce119c4d624defab4"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">decoderPerRequest</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="k"><span class="pre">false</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession6Config17decoderPerRequestE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession6Config13cudaGraphModeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession6Config13cudaGraphModeE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession6Config13cudaGraphModeE"></span><span id="tensorrt_llm::runtime::GptSession::Config::cudaGraphMode__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1Config_1a2584bcfc17050c7ce1370868ccb1e9b9"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">cudaGraphMode</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="k"><span class="pre">false</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession6Config13cudaGraphModeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession6Config13kvCacheConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession6Config13kvCacheConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession6Config13kvCacheConfigE"></span><span id="tensorrt_llm::runtime::GptSession::Config::kvCacheConfig__KvCacheConfig"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1Config_1ab6b5dafa19a5fa0c82d979e17d0a9b0f"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession13KvCacheConfigE" title="tensorrt_llm::runtime::GptSession::KvCacheConfig"><span class="n"><span class="pre">KvCacheConfig</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kvCacheConfig</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession6Config13kvCacheConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession6Config17ctxMicroBatchSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession6Config17ctxMicroBatchSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession6Config17ctxMicroBatchSizeE"></span><span id="tensorrt_llm::runtime::GptSession::Config::ctxMicroBatchSize__std::optional:SizeType32:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1Config_1a407b6f91751b5c28a412390e9a3bf09c"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">ctxMicroBatchSize</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">nullopt</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession6Config17ctxMicroBatchSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession6Config17genMicroBatchSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession6Config17genMicroBatchSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession6Config17genMicroBatchSizeE"></span><span id="tensorrt_llm::runtime::GptSession::Config::genMicroBatchSize__std::optional:SizeType32:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1Config_1ad04f7614d9bf5dab8476a374837b4571"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">genMicroBatchSize</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">nullopt</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession6Config17genMicroBatchSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession6Config12decodingModeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession6Config12decodingModeE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession6Config12decodingModeE"></span><span id="tensorrt_llm::runtime::GptSession::Config::decodingMode__std::optional:executor::DecodingMode:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1Config_1a0d117ec1fda29b76ca9926c678c5c97a"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executorE" title="tensorrt_llm::executor"><span class="n"><span class="pre">executor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executor12DecodingModeE" title="tensorrt_llm::executor::DecodingMode"><span class="n"><span class="pre">DecodingMode</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">decodingMode</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">nullopt</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession6Config12decodingModeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession6Config17normalizeLogProbsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession6Config17normalizeLogProbsE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession6Config17normalizeLogProbsE"></span><span id="tensorrt_llm::runtime::GptSession::Config::normalizeLogProbs__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1Config_1a3f3f437a1db80088044d47ac77d6ccea"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">normalizeLogProbs</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">true</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession6Config17normalizeLogProbsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession17CudaGraphExecutorE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession17CudaGraphExecutorE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession17CudaGraphExecutorE"></span><span id="tensorrt_llm::runtime::GptSession::CudaGraphExecutor"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1CudaGraphExecutor"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">CudaGraphExecutor</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession17CudaGraphExecutorE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor17CudaGraphExecutorEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor17CudaGraphExecutorEv"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor17CudaGraphExecutorEv"></span><span id="tensorrt_llm::runtime::GptSession::CudaGraphExecutor::CudaGraphExecutor"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1CudaGraphExecutor_1a0f76e8397ca3097bc3451e2251c4d6bd"></span><span class="sig-name descname"><span class="n"><span class="pre">CudaGraphExecutor</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">default</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor17CudaGraphExecutorEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession17CudaGraphExecutorD0Ev">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession17CudaGraphExecutorD0Ev"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession17CudaGraphExecutorD0Ev"></span><span id="tensorrt_llm::runtime::GptSession::CudaGraphExecutor::~CudaGraphExecutor"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1CudaGraphExecutor_1a0d609c019d81064a9cdf235352d7c25f"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">~CudaGraphExecutor</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession17CudaGraphExecutorD0Ev" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor11hasInstanceEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor11hasInstanceEv"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor11hasInstanceEv"></span><span id="tensorrt_llm::runtime::GptSession::CudaGraphExecutor::hasInstance"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1CudaGraphExecutor_1a11ce3f651b7e0c1dbba82460a85cfd3b"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">hasInstance</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor11hasInstanceEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor5clearEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor5clearEv"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor5clearEv"></span><span id="tensorrt_llm::runtime::GptSession::CudaGraphExecutor::clear"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1CudaGraphExecutor_1a6bd0bf9ff37bb89cb06d03fb19b2de59"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">clear</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor5clearEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor16prepareNextGraphERK11TllmRuntime10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor16prepareNextGraphERK11TllmRuntime10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor16prepareNextGraphERK11TllmRuntime10SizeType32"></span><span id="tensorrt_llm::runtime::GptSession::CudaGraphExecutor::prepareNextGraph__TllmRuntimeCR.SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1CudaGraphExecutor_1ab785759fb842eb024f71538d57024dd7"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">prepareNextGraph</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">TllmRuntime</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">runtime</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">nextContextId</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor16prepareNextGraphERK11TllmRuntime10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor6launchERK10CudaStream">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor6launchERK10CudaStream"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor6launchERK10CudaStream"></span><span id="tensorrt_llm::runtime::GptSession::CudaGraphExecutor::launch__CudaStreamCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1CudaGraphExecutor_1a9d32b4b5046b27364c3e265b433b62a4"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">launch</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10CudaStreamE" title="tensorrt_llm::runtime::CudaStream"><span class="n"><span class="pre">CudaStream</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">stream</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor6launchERK10CudaStream" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-functions">Private Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor6createERK11cudaGraph_t">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor6createERK11cudaGraph_t"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor6createERK11cudaGraph_t"></span><span id="tensorrt_llm::runtime::GptSession::CudaGraphExecutor::create__cudaGraph_tCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1CudaGraphExecutor_1a368e5d334bee7c4afd9085fdbe98f02e"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">create</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">cudaGraph_t</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">graph</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor6createERK11cudaGraph_t" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor6updateERK11cudaGraph_t">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor6updateERK11cudaGraph_t"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor6updateERK11cudaGraph_t"></span><span id="tensorrt_llm::runtime::GptSession::CudaGraphExecutor::update__cudaGraph_tCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1CudaGraphExecutor_1a776709d610ef7122a55d1a9dd211acb9"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">update</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">cudaGraph_t</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">graph</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor6updateERK11cudaGraph_t" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor14uploadToStreamERK10CudaStream">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor14uploadToStreamERK10CudaStream"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor14uploadToStreamERK10CudaStream"></span><span id="tensorrt_llm::runtime::GptSession::CudaGraphExecutor::uploadToStream__CudaStreamCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1CudaGraphExecutor_1af2cd5b05bdcf087f91696bc83daa90e0"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">uploadToStream</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10CudaStreamE" title="tensorrt_llm::runtime::CudaStream"><span class="n"><span class="pre">CudaStream</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">stream</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor14uploadToStreamERK10CudaStream" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-members">Private Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor9mInstanceE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor9mInstanceE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor9mInstanceE"></span><span id="tensorrt_llm::runtime::GptSession::CudaGraphExecutor::mInstance__cudaGraphExec_t"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1CudaGraphExecutor_1ac8e98428b96096b5509dc101b66be172"></span><span class="n"><span class="pre">cudaGraphExec_t</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mInstance</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor9mInstanceE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession18GenerationProfilerE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession18GenerationProfilerE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession18GenerationProfilerE"></span><span id="tensorrt_llm::runtime::GptSession::GenerationProfiler"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1GenerationProfiler"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">GenerationProfiler</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession18GenerationProfilerE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="docutils container">
|
||
<em>#include <gptSession.h></em></div>
|
||
<p>Optional profiler class to profile the generation phase of an inference request. </p>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession18GenerationProfiler18GenerationProfilerEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession18GenerationProfiler18GenerationProfilerEv"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession18GenerationProfiler18GenerationProfilerEv"></span><span id="tensorrt_llm::runtime::GptSession::GenerationProfiler::GenerationProfiler"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1GenerationProfiler_1a84c8d099dc4d6db6d9cbf74e2549eab7"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">GenerationProfiler</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession18GenerationProfiler18GenerationProfilerEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10GptSession18GenerationProfiler8getStartEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10GptSession18GenerationProfiler8getStartEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime10GptSession18GenerationProfiler8getStartEv"></span><span id="tensorrt_llm::runtime::GptSession::GenerationProfiler::getStartC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1GenerationProfiler_1a18b141bd35f3e01f590bbfd873500aa1"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEventE" title="tensorrt_llm::runtime::CudaEvent"><span class="n"><span class="pre">CudaEvent</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="n"><span class="pre">getStart</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession18GenerationProfiler8getStartEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10GptSession18GenerationProfiler6getEndEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10GptSession18GenerationProfiler6getEndEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime10GptSession18GenerationProfiler6getEndEv"></span><span id="tensorrt_llm::runtime::GptSession::GenerationProfiler::getEndC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1GenerationProfiler_1a1800846dabb664f120e0a1aa430784f2"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEventE" title="tensorrt_llm::runtime::CudaEvent"><span class="n"><span class="pre">CudaEvent</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="n"><span class="pre">getEnd</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession18GenerationProfiler6getEndEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession18GenerationProfiler16getElapsedTimeMsEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession18GenerationProfiler16getElapsedTimeMsEv"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession18GenerationProfiler16getElapsedTimeMsEv"></span><span id="tensorrt_llm::runtime::GptSession::GenerationProfiler::getElapsedTimeMs"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1GenerationProfiler_1aa0ce46188eeaf923955ada55724bb3bc"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">float</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getElapsedTimeMs</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession18GenerationProfiler16getElapsedTimeMsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession18GenerationProfiler5flagsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession18GenerationProfiler5flagsE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession18GenerationProfiler5flagsE"></span><span id="tensorrt_llm::runtime::GptSession::GenerationProfiler::flags__unsigned-i"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1GenerationProfiler_1ae4cffa04fd70e93d3c44d46370f1c7d8"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="kt"><span class="pre">unsigned</span></span><span class="w"> </span><span class="kt"><span class="pre">int</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">flags</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="n"><span class="pre">cudaEventDefault</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession18GenerationProfiler5flagsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-members">Private Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession18GenerationProfiler5startE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession18GenerationProfiler5startE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession18GenerationProfiler5startE"></span><span id="tensorrt_llm::runtime::GptSession::GenerationProfiler::start__CudaEvent"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1GenerationProfiler_1a145183c372a872e41c0cf6ff54e7decf"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEventE" title="tensorrt_llm::runtime::CudaEvent"><span class="n"><span class="pre">CudaEvent</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">start</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession18GenerationProfiler5startE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession18GenerationProfiler3endE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession18GenerationProfiler3endE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession18GenerationProfiler3endE"></span><span id="tensorrt_llm::runtime::GptSession::GenerationProfiler::end__CudaEvent"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1GenerationProfiler_1a2ddc664abe8247bfdafd6ad96d2d5ada"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEventE" title="tensorrt_llm::runtime::CudaEvent"><span class="n"><span class="pre">CudaEvent</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">end</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession18GenerationProfiler3endE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession16MicroBatchConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession16MicroBatchConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession16MicroBatchConfigE"></span><span id="tensorrt_llm::runtime::GptSession::MicroBatchConfig"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1MicroBatchConfig"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">MicroBatchConfig</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession16MicroBatchConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession16MicroBatchConfig16MicroBatchConfigEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession16MicroBatchConfig16MicroBatchConfigEv"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession16MicroBatchConfig16MicroBatchConfigEv"></span><span id="tensorrt_llm::runtime::GptSession::MicroBatchConfig::MicroBatchConfig"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1MicroBatchConfig_1a1eac3f1b519be9f762d23157d6b04444"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">MicroBatchConfig</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession16MicroBatchConfig16MicroBatchConfigEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession16MicroBatchConfig16MicroBatchConfigE10SizeType3210SizeType32NSt8optionalI10SizeType32EENSt8optionalI10SizeType32EE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession16MicroBatchConfig16MicroBatchConfigE10SizeType3210SizeType32NSt8optionalI10SizeType32EENSt8optionalI10SizeType32EE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession16MicroBatchConfig16MicroBatchConfigE10SizeType3210SizeType32NSt8optionalI10SizeType32EENSt8optionalI10SizeType32EE"></span><span id="tensorrt_llm::runtime::GptSession::MicroBatchConfig::MicroBatchConfig__SizeType32.SizeType32.std::optional:SizeType32:.std::optional:SizeType32:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1MicroBatchConfig_1a1f3dd337346879404943a255f75634d6"></span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">MicroBatchConfig</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxBatchSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">pipelineParallelism</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="n sig-param"><span class="pre">genMicroBatchSize</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="n sig-param"><span class="pre">ctxMicroBatchSize</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession16MicroBatchConfig16MicroBatchConfigE10SizeType3210SizeType32NSt8optionalI10SizeType32EENSt8optionalI10SizeType32EE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10GptSession16MicroBatchConfig12numCtxPerGenEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10GptSession16MicroBatchConfig12numCtxPerGenEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime10GptSession16MicroBatchConfig12numCtxPerGenEv"></span><span id="tensorrt_llm::runtime::GptSession::MicroBatchConfig::numCtxPerGenCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1MicroBatchConfig_1a017e309e19941576decf04e278657c52"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">numCtxPerGen</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession16MicroBatchConfig12numCtxPerGenEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10GptSession16MicroBatchConfig13getGenGraphIdE10SizeType3210SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10GptSession16MicroBatchConfig13getGenGraphIdE10SizeType3210SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime10GptSession16MicroBatchConfig13getGenGraphIdE10SizeType3210SizeType32"></span><span id="tensorrt_llm::runtime::GptSession::MicroBatchConfig::getGenGraphId__SizeType32.SizeType32CCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1MicroBatchConfig_1a44c6d6d7ad24de1e136d65bb52bf6e12"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getGenGraphId</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">flipFlopId</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">generationBatchId</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime10GptSession16MicroBatchConfig13getGenGraphIdE10SizeType3210SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>flip-flop between 2 graph instances for each generation batch. </p>
|
||
</dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession16MicroBatchConfig13numCtxBatchesE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession16MicroBatchConfig13numCtxBatchesE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession16MicroBatchConfig13numCtxBatchesE"></span><span id="tensorrt_llm::runtime::GptSession::MicroBatchConfig::numCtxBatches__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1MicroBatchConfig_1a74d1e38dd3e66a1b37d46534ef778f83"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">numCtxBatches</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession16MicroBatchConfig13numCtxBatchesE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession16MicroBatchConfig13numGenBatchesE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession16MicroBatchConfig13numGenBatchesE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession16MicroBatchConfig13numGenBatchesE"></span><span id="tensorrt_llm::runtime::GptSession::MicroBatchConfig::numGenBatches__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1MicroBatchConfig_1adf01eb8c15b052c2cc9fa94dba54b5b2"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">numGenBatches</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession16MicroBatchConfig13numGenBatchesE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession16MicroBatchConfig12ctxBatchSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession16MicroBatchConfig12ctxBatchSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession16MicroBatchConfig12ctxBatchSizeE"></span><span id="tensorrt_llm::runtime::GptSession::MicroBatchConfig::ctxBatchSize__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1MicroBatchConfig_1a1c6ba1cb67c250aea112c41c70125004"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">ctxBatchSize</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession16MicroBatchConfig12ctxBatchSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession16MicroBatchConfig12genBatchSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession16MicroBatchConfig12genBatchSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession16MicroBatchConfig12genBatchSizeE"></span><span id="tensorrt_llm::runtime::GptSession::MicroBatchConfig::genBatchSize__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1MicroBatchConfig_1aac3178a24a206e3750adbc47e1dc2ead"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">genBatchSize</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession16MicroBatchConfig12genBatchSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime5utilsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime5utilsE"></span><span id="_CPPv2N12tensorrt_llm7runtime5utilsE"></span><span id="tensorrt_llm::runtime::utils"></span><span class="target" id="namespacetensorrt__llm_1_1runtime_1_1utils"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">utils</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime5utilsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-functions">Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime5utils10loadEngineERKNSt6stringE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime5utils10loadEngineERKNSt6stringE"></span><span id="_CPPv2N12tensorrt_llm7runtime5utils10loadEngineERKNSt6stringE"></span><span id="tensorrt_llm::runtime::utils::loadEngine__ssCR"></span><span class="target" id="gptSession_8h_1a6116ef7abb02bc99e9119a49e0fb4053"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">uint8_t</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">loadEngine</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">enginePath</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime5utils10loadEngineERKNSt6stringE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="ibuffer-h">
|
||
<h2>iBuffer.h<a class="headerlink" href="#ibuffer-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kGPUEEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kGPUEEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kGPUEEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1MemoryTypeString_3_01MemoryType_1_1kGPU_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">MemoryTypeString</span></span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="tensorrt_llm::runtime::MemoryType"><span class="n"><span class="pre">MemoryType</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryType4kGPUE" title="tensorrt_llm::runtime::MemoryType::kGPU"><span class="n"><span class="pre">kGPU</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4IEN12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kGPUEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kGPUEE5valueE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kGPUEE5valueE"></span><span id="_CPPv2N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kGPUEE5valueE"></span><span id="tensorrt_llm::runtime::MemoryTypeString:MemoryType::kGPU:::value__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1MemoryTypeString_3_01MemoryType_1_1kGPU_01_4_1a0fc1c309e17dd52d719cafddbcc6cb12"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"GPU"</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kGPUEE5valueE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kCPUEEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kCPUEEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kCPUEEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1MemoryTypeString_3_01MemoryType_1_1kCPU_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">MemoryTypeString</span></span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="tensorrt_llm::runtime::MemoryType"><span class="n"><span class="pre">MemoryType</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryType4kCPUE" title="tensorrt_llm::runtime::MemoryType::kCPU"><span class="n"><span class="pre">kCPU</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4IEN12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kCPUEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kCPUEE5valueE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kCPUEE5valueE"></span><span id="_CPPv2N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kCPUEE5valueE"></span><span id="tensorrt_llm::runtime::MemoryTypeString:MemoryType::kCPU:::value__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1MemoryTypeString_3_01MemoryType_1_1kCPU_01_4_1af3b289e3cfd246c2b1ee25aec5eeb6e1"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"CPU"</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kCPUEE5valueE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType7kPINNEDEEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType7kPINNEDEEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType7kPINNEDEEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1MemoryTypeString_3_01MemoryType_1_1kPINNED_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">MemoryTypeString</span></span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="tensorrt_llm::runtime::MemoryType"><span class="n"><span class="pre">MemoryType</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryType7kPINNEDE" title="tensorrt_llm::runtime::MemoryType::kPINNED"><span class="n"><span class="pre">kPINNED</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4IEN12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType7kPINNEDEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType7kPINNEDEE5valueE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType7kPINNEDEE5valueE"></span><span id="_CPPv2N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType7kPINNEDEE5valueE"></span><span id="tensorrt_llm::runtime::MemoryTypeString:MemoryType::kPINNED:::value__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1MemoryTypeString_3_01MemoryType_1_1kPINNED_01_4_1a5e877eb90ca7cbd5b5269d4f092e69e4"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"PINNED"</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType7kPINNEDEE5valueE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kUVMEEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kUVMEEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kUVMEEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1MemoryTypeString_3_01MemoryType_1_1kUVM_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">MemoryTypeString</span></span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="tensorrt_llm::runtime::MemoryType"><span class="n"><span class="pre">MemoryType</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryType4kUVME" title="tensorrt_llm::runtime::MemoryType::kUVM"><span class="n"><span class="pre">kUVM</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4IEN12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kUVMEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kUVMEE5valueE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kUVMEE5valueE"></span><span id="_CPPv2N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kUVMEE5valueE"></span><span id="tensorrt_llm::runtime::MemoryTypeString:MemoryType::kUVM:::value__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1MemoryTypeString_3_01MemoryType_1_1kUVM_01_4_1a743212e7128317c3b6a0af4ad619afa9"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"UVM"</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kUVMEE5valueE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType11kPINNEDPOOLEEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType11kPINNEDPOOLEEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType11kPINNEDPOOLEEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1MemoryTypeString_3_01MemoryType_1_1kPINNEDPOOL_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">MemoryTypeString</span></span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="tensorrt_llm::runtime::MemoryType"><span class="n"><span class="pre">MemoryType</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryType11kPINNEDPOOLE" title="tensorrt_llm::runtime::MemoryType::kPINNEDPOOL"><span class="n"><span class="pre">kPINNEDPOOL</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4IEN12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType11kPINNEDPOOLEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType11kPINNEDPOOLEE5valueE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType11kPINNEDPOOLEE5valueE"></span><span id="_CPPv2N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType11kPINNEDPOOLEE5valueE"></span><span id="tensorrt_llm::runtime::MemoryTypeString:MemoryType::kPINNEDPOOL:::value__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1MemoryTypeString_3_01MemoryType_1_1kPINNEDPOOL_01_4_1aad3eee9ea444088033995a877a00e9c9"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"PINNEDPOOL"</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType11kPINNEDPOOLEE5valueE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kFLOATEEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kFLOATEEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kFLOATEEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kFLOAT_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">DataTypeTraits</span></span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">kFLOAT</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kFLOATEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kFLOATEE4typeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kFLOATEE4typeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kFLOATEE4typeE"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kFLOAT_01_4_1af6ef0c2bf073670006e1a68fd861495b"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">type</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="kt"><span class="pre">float</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kFLOATEE4typeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kFLOATEE4nameE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kFLOATEE4nameE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kFLOATEE4nameE"></span><span id="tensorrt_llm::runtime::DataTypeTraits:nvinfer1::DataType::kFLOAT:::name__cA"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kFLOAT_01_4_1a7253109d1e4c01ed33157460dba35ec7"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">char</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">name</span></span></span><span class="p"><span class="pre">[</span></span><span class="p"><span class="pre">]</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"float"</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kFLOATEE4nameE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kFLOATEE4sizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kFLOATEE4sizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kFLOATEE4sizeE"></span><span id="tensorrt_llm::runtime::DataTypeTraits:nvinfer1::DataType::kFLOAT:::size__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kFLOAT_01_4_1afff1a8fde1dc21c665c8e1f3f8cc3fbd"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">size</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">sizeof</span></span><span class="p"><span class="pre">(</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kFLOATEE4typeE" title="tensorrt_llm::runtime::DataTypeTraits<nvinfer1::DataType::kFLOAT>::type"><span class="n"><span class="pre">type</span></span></a><span class="p"><span class="pre">)</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kFLOATEE4sizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kHALFEEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kHALFEEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kHALFEEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kHALF_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">DataTypeTraits</span></span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">kHALF</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kHALFEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kHALFEE4typeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kHALFEE4typeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kHALFEE4typeE"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kHALF_01_4_1a08c9427d16066ff8ee9e9deb35a91871"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">type</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">half</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kHALFEE4typeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kHALFEE4nameE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kHALFEE4nameE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kHALFEE4nameE"></span><span id="tensorrt_llm::runtime::DataTypeTraits:nvinfer1::DataType::kHALF:::name__cA"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kHALF_01_4_1a661db84e273578749fbd43d1f6829aa0"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">char</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">name</span></span></span><span class="p"><span class="pre">[</span></span><span class="p"><span class="pre">]</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"half"</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kHALFEE4nameE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kHALFEE4sizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kHALFEE4sizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kHALFEE4sizeE"></span><span id="tensorrt_llm::runtime::DataTypeTraits:nvinfer1::DataType::kHALF:::size__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kHALF_01_4_1aad4af9eb10a14553b0dc46bd210454c1"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">size</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">sizeof</span></span><span class="p"><span class="pre">(</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kHALFEE4typeE" title="tensorrt_llm::runtime::DataTypeTraits<nvinfer1::DataType::kHALF>::type"><span class="n"><span class="pre">type</span></span></a><span class="p"><span class="pre">)</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kHALFEE4sizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kINT8EEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kINT8EEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kINT8EEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT8_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">DataTypeTraits</span></span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">kINT8</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kINT8EEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kINT8EE4typeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kINT8EE4typeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kINT8EE4typeE"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT8_01_4_1a51b8ee14d86b4b2a6cc5e57a1d524c43"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">type</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">int8_t</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kINT8EE4typeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kINT8EE4nameE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kINT8EE4nameE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kINT8EE4nameE"></span><span id="tensorrt_llm::runtime::DataTypeTraits:nvinfer1::DataType::kINT8:::name__cA"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT8_01_4_1ad45ffa244d70cff61abcc7c89436e839"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">char</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">name</span></span></span><span class="p"><span class="pre">[</span></span><span class="p"><span class="pre">]</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"int8"</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kINT8EE4nameE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kINT8EE4sizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kINT8EE4sizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kINT8EE4sizeE"></span><span id="tensorrt_llm::runtime::DataTypeTraits:nvinfer1::DataType::kINT8:::size__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT8_01_4_1ad29e96b3fe577dba8524735a4f42f768"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">size</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">sizeof</span></span><span class="p"><span class="pre">(</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kINT8EE4typeE" title="tensorrt_llm::runtime::DataTypeTraits<nvinfer1::DataType::kINT8>::type"><span class="n"><span class="pre">type</span></span></a><span class="p"><span class="pre">)</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kINT8EE4sizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT32_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">DataTypeTraits</span></span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">kINT32</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EE4typeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EE4typeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EE4typeE"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT32_01_4_1a37fa1b65a6bece3eb638f7693132053c"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">type</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">int32_t</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EE4typeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EE4nameE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EE4nameE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EE4nameE"></span><span id="tensorrt_llm::runtime::DataTypeTraits:nvinfer1::DataType::kINT32:::name__cA"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT32_01_4_1a98f4973f56a92c96b5bd1c4298b4ca72"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">char</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">name</span></span></span><span class="p"><span class="pre">[</span></span><span class="p"><span class="pre">]</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"int32"</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EE4nameE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EE4sizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EE4sizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EE4sizeE"></span><span id="tensorrt_llm::runtime::DataTypeTraits:nvinfer1::DataType::kINT32:::size__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT32_01_4_1ad02b1cf34aed9754967e7bb76a55fb0b"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">size</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">sizeof</span></span><span class="p"><span class="pre">(</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EE4typeE" title="tensorrt_llm::runtime::DataTypeTraits<nvinfer1::DataType::kINT32>::type"><span class="n"><span class="pre">type</span></span></a><span class="p"><span class="pre">)</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EE4sizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT64_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">DataTypeTraits</span></span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">kINT64</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EE4typeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EE4typeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EE4typeE"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT64_01_4_1a288b116cbff3532cf2b80e7206277846"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">type</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">int64_t</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EE4typeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EE4nameE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EE4nameE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EE4nameE"></span><span id="tensorrt_llm::runtime::DataTypeTraits:nvinfer1::DataType::kINT64:::name__cA"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT64_01_4_1a912e86b1f6bb4572292b27ed87819384"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">char</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">name</span></span></span><span class="p"><span class="pre">[</span></span><span class="p"><span class="pre">]</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"int64"</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EE4nameE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EE4sizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EE4sizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EE4sizeE"></span><span id="tensorrt_llm::runtime::DataTypeTraits:nvinfer1::DataType::kINT64:::size__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT64_01_4_1ac9994f6271fd3dd86e3be01df67a5577"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">size</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">sizeof</span></span><span class="p"><span class="pre">(</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EE4typeE" title="tensorrt_llm::runtime::DataTypeTraits<nvinfer1::DataType::kINT64>::type"><span class="n"><span class="pre">type</span></span></a><span class="p"><span class="pre">)</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EE4sizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EXL1EEEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EXL1EEEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EXtrueEEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT32_00_01true_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">DataTypeTraits</span></span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">kINT32</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="k"><span class="pre">true</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EXL1EEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EXL1EEE4typeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EXL1EEE4typeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EXtrueEE4typeE"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT32_00_01true_01_4_1a41a0f3bcdb82b843a9594fbe518c148d"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">type</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">uint32_t</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EXL1EEE4typeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EXL1EEE4nameE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EXL1EEE4nameE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EXtrueEE4nameE"></span><span id="tensorrt_llm::runtime::DataTypeTraits:nvinfer1::DataType::kINT32.true:::name__cA"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT32_00_01true_01_4_1a455a075a88a89b0bf29b5ac1afc54320"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">char</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">name</span></span></span><span class="p"><span class="pre">[</span></span><span class="p"><span class="pre">]</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"uint32"</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EXL1EEE4nameE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EXL1EEE4sizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EXL1EEE4sizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EXtrueEE4sizeE"></span><span id="tensorrt_llm::runtime::DataTypeTraits:nvinfer1::DataType::kINT32.true:::size__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT32_00_01true_01_4_1a420375ac7c713ee26156c7e143327393"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">size</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">sizeof</span></span><span class="p"><span class="pre">(</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EXL1EEE4typeE" title="tensorrt_llm::runtime::DataTypeTraits<nvinfer1::DataType::kINT32, true>::type"><span class="n"><span class="pre">type</span></span></a><span class="p"><span class="pre">)</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT32EXL1EEE4sizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EXL1EEEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EXL1EEEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EXtrueEEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT64_00_01true_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">DataTypeTraits</span></span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">kINT64</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="k"><span class="pre">true</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4IEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EXL1EEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EXL1EEE4typeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EXL1EEE4typeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EXtrueEE4typeE"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT64_00_01true_01_4_1a8376cfb6029bfec1e7aaa9c29369b094"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">type</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">uint64_t</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EXL1EEE4typeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EXL1EEE4nameE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EXL1EEE4nameE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EXtrueEE4nameE"></span><span id="tensorrt_llm::runtime::DataTypeTraits:nvinfer1::DataType::kINT64.true:::name__cA"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT64_00_01true_01_4_1ae082c5ad1c2c409b8cbedc64b69e7f1d"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">char</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">name</span></span></span><span class="p"><span class="pre">[</span></span><span class="p"><span class="pre">]</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"uint64"</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EXL1EEE4nameE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EXL1EEE4sizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EXL1EEE4sizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EXtrueEE4sizeE"></span><span id="tensorrt_llm::runtime::DataTypeTraits:nvinfer1::DataType::kINT64.true:::size__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT64_00_01true_01_4_1af5140fb67b9b5cae4c40ea6f998f1223"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">size</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">sizeof</span></span><span class="p"><span class="pre">(</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EXL1EEE4typeE" title="tensorrt_llm::runtime::DataTypeTraits<nvinfer1::DataType::kINT64, true>::type"><span class="n"><span class="pre">type</span></span></a><span class="p"><span class="pre">)</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kINT64EXL1EEE4sizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I_bEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kBOOLE9kUnsignedEE">
|
||
<span id="_CPPv3I_bEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kBOOLE9kUnsignedEE"></span><span id="_CPPv2I_bEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kBOOLE9kUnsignedEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kUnsigned</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kBOOL_00_01kUnsigned_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">DataTypeTraits</span></span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">kBOOL</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4I_bEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kBOOLE9kUnsignedEE" title="tensorrt_llm::runtime::DataTypeTraits<nvinfer1::DataType::kBOOL, kUnsigned>::kUnsigned"><span class="n"><span class="pre">kUnsigned</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4I_bEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kBOOLE9kUnsignedEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kBOOLE9kUnsignedE4typeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kBOOLE9kUnsignedE4typeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kBOOLE9kUnsignedE4typeE"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kBOOL_00_01kUnsigned_01_4_1a3ef7ec1c22f6cc0b397ab5eb28c232eb"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">type</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kBOOLE9kUnsignedE4typeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kBOOLE9kUnsignedE4nameE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kBOOLE9kUnsignedE4nameE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kBOOLE9kUnsignedE4nameE"></span><span id="tensorrt_llm::runtime::DataTypeTraits:nvinfer1::DataType::kBOOL.kUnsigned:::name__cA"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kBOOL_00_01kUnsigned_01_4_1a60ce7e6d8364224764d2415e4b709925"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">char</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">name</span></span></span><span class="p"><span class="pre">[</span></span><span class="p"><span class="pre">]</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"bool"</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kBOOLE9kUnsignedE4nameE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kBOOLE9kUnsignedE4sizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kBOOLE9kUnsignedE4sizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kBOOLE9kUnsignedE4sizeE"></span><span id="tensorrt_llm::runtime::DataTypeTraits:nvinfer1::DataType::kBOOL.kUnsigned:::size__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kBOOL_00_01kUnsigned_01_4_1aea0d24ee412e727b1e7d1831501600df"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">size</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">sizeof</span></span><span class="p"><span class="pre">(</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kBOOLE9kUnsignedE4typeE" title="tensorrt_llm::runtime::DataTypeTraits<nvinfer1::DataType::kBOOL, kUnsigned>::type"><span class="n"><span class="pre">type</span></span></a><span class="p"><span class="pre">)</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType5kBOOLE9kUnsignedE4sizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I_bEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kUINT8E9kUnsignedEE">
|
||
<span id="_CPPv3I_bEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kUINT8E9kUnsignedEE"></span><span id="_CPPv2I_bEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kUINT8E9kUnsignedEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kUnsigned</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kUINT8_00_01kUnsigned_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">DataTypeTraits</span></span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">kUINT8</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4I_bEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kUINT8E9kUnsignedEE" title="tensorrt_llm::runtime::DataTypeTraits<nvinfer1::DataType::kUINT8, kUnsigned>::kUnsigned"><span class="n"><span class="pre">kUnsigned</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4I_bEN12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kUINT8E9kUnsignedEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kUINT8E9kUnsignedE4typeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kUINT8E9kUnsignedE4typeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kUINT8E9kUnsignedE4typeE"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kUINT8_00_01kUnsigned_01_4_1a5a3089fdad68ca4ce8bf36137d18a147"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">type</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">uint8_t</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kUINT8E9kUnsignedE4typeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kUINT8E9kUnsignedE4nameE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kUINT8E9kUnsignedE4nameE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kUINT8E9kUnsignedE4nameE"></span><span id="tensorrt_llm::runtime::DataTypeTraits:nvinfer1::DataType::kUINT8.kUnsigned:::name__cA"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kUINT8_00_01kUnsigned_01_4_1a997c0219602be3dfcde787c95462e1d5"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">char</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">name</span></span></span><span class="p"><span class="pre">[</span></span><span class="p"><span class="pre">]</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"uint8"</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kUINT8E9kUnsignedE4nameE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kUINT8E9kUnsignedE4sizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kUINT8E9kUnsignedE4sizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kUINT8E9kUnsignedE4sizeE"></span><span id="tensorrt_llm::runtime::DataTypeTraits:nvinfer1::DataType::kUINT8.kUnsigned:::size__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kUINT8_00_01kUnsigned_01_4_1abb3870f3b3387d0cfedb4af98f5d71a6"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">size</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">sizeof</span></span><span class="p"><span class="pre">(</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kUINT8E9kUnsignedE4typeE" title="tensorrt_llm::runtime::DataTypeTraits<nvinfer1::DataType::kUINT8, kUnsigned>::type"><span class="n"><span class="pre">type</span></span></a><span class="p"><span class="pre">)</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsIN8nvinfer18DataType6kUINT8E9kUnsignedE4sizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeINSt6int8_tEEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime11TRTDataTypeINSt6int8_tEEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime11TRTDataTypeINSt6int8_tEEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01std_1_1int8__t_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TRTDataType</span></span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">int8_t</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeINSt6int8_tEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11TRTDataTypeINSt6int8_tEE5valueE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11TRTDataTypeINSt6int8_tEE5valueE"></span><span id="_CPPv2N12tensorrt_llm7runtime11TRTDataTypeINSt6int8_tEE5valueE"></span><span id="tensorrt_llm::runtime::TRTDataType:std::int8_t:::value__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01std_1_1int8__t_01_4_1a996ff7940394a814ae164728312bf5ff"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">kINT8</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeINSt6int8_tEE5valueE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeINSt7int32_tEEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime11TRTDataTypeINSt7int32_tEEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime11TRTDataTypeINSt7int32_tEEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01std_1_1int32__t_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TRTDataType</span></span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">int32_t</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeINSt7int32_tEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11TRTDataTypeINSt7int32_tEE5valueE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11TRTDataTypeINSt7int32_tEE5valueE"></span><span id="_CPPv2N12tensorrt_llm7runtime11TRTDataTypeINSt7int32_tEE5valueE"></span><span id="tensorrt_llm::runtime::TRTDataType:std::int32_t:::value__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01std_1_1int32__t_01_4_1a10403a956a1d979e0cd10c8034e9f3e1"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">kINT32</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeINSt7int32_tEE5valueE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeINSt8uint32_tEEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime11TRTDataTypeINSt8uint32_tEEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime11TRTDataTypeINSt8uint32_tEEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01std_1_1uint32__t_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TRTDataType</span></span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">uint32_t</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeINSt8uint32_tEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11TRTDataTypeINSt8uint32_tEE5valueE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11TRTDataTypeINSt8uint32_tEE5valueE"></span><span id="_CPPv2N12tensorrt_llm7runtime11TRTDataTypeINSt8uint32_tEE5valueE"></span><span id="tensorrt_llm::runtime::TRTDataType:std::uint32_t:::value__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01std_1_1uint32__t_01_4_1a96b97cf0edb4ad984dd69278acfabd10"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14BufferDataTypeE" title="tensorrt_llm::runtime::BufferDataType"><span class="n"><span class="pre">BufferDataType</span></span></a><span class="p"><span class="pre">{</span></span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">kINT32</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="k"><span class="pre">true</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeINSt8uint32_tEE5valueE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeINSt7int64_tEEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime11TRTDataTypeINSt7int64_tEEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime11TRTDataTypeINSt7int64_tEEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01std_1_1int64__t_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TRTDataType</span></span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">int64_t</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeINSt7int64_tEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11TRTDataTypeINSt7int64_tEE5valueE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11TRTDataTypeINSt7int64_tEE5valueE"></span><span id="_CPPv2N12tensorrt_llm7runtime11TRTDataTypeINSt7int64_tEE5valueE"></span><span id="tensorrt_llm::runtime::TRTDataType:std::int64_t:::value__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01std_1_1int64__t_01_4_1af2839adf78453826b3a7f8422c9130fb"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">kINT64</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeINSt7int64_tEE5valueE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeINSt8uint64_tEEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime11TRTDataTypeINSt8uint64_tEEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime11TRTDataTypeINSt8uint64_tEEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01std_1_1uint64__t_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TRTDataType</span></span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">uint64_t</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeINSt8uint64_tEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11TRTDataTypeINSt8uint64_tEE5valueE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11TRTDataTypeINSt8uint64_tEE5valueE"></span><span id="_CPPv2N12tensorrt_llm7runtime11TRTDataTypeINSt8uint64_tEE5valueE"></span><span id="tensorrt_llm::runtime::TRTDataType:std::uint64_t:::value__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01std_1_1uint64__t_01_4_1a6d279398fbfedbbfffd0f7fbd1eb80b8"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14BufferDataTypeE" title="tensorrt_llm::runtime::BufferDataType"><span class="n"><span class="pre">BufferDataType</span></span></a><span class="p"><span class="pre">{</span></span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">kINT64</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="k"><span class="pre">true</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeINSt8uint64_tEE5valueE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeINSt7uint8_tEEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime11TRTDataTypeINSt7uint8_tEEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime11TRTDataTypeINSt7uint8_tEEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01std_1_1uint8__t_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TRTDataType</span></span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">uint8_t</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeINSt7uint8_tEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11TRTDataTypeINSt7uint8_tEE5valueE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11TRTDataTypeINSt7uint8_tEE5valueE"></span><span id="_CPPv2N12tensorrt_llm7runtime11TRTDataTypeINSt7uint8_tEE5valueE"></span><span id="tensorrt_llm::runtime::TRTDataType:std::uint8_t:::value__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01std_1_1uint8__t_01_4_1a60d418bc6cea9f845ee353d78f0c3d66"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">kUINT8</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeINSt7uint8_tEE5valueE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeIN7kernels12KVCacheIndexEEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime11TRTDataTypeIN7kernels12KVCacheIndexEEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime11TRTDataTypeIN7kernels12KVCacheIndexEEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01kernels_1_1KVCacheIndex_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TRTDataType</span></span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">kernels</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">KVCacheIndex</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeIN7kernels12KVCacheIndexEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11TRTDataTypeIN7kernels12KVCacheIndexEE5valueE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11TRTDataTypeIN7kernels12KVCacheIndexEE5valueE"></span><span id="_CPPv2N12tensorrt_llm7runtime11TRTDataTypeIN7kernels12KVCacheIndexEE5valueE"></span><span id="tensorrt_llm::runtime::TRTDataType:kernels::KVCacheIndex:::value__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01kernels_1_1KVCacheIndex_01_4_1a8a9684f587b9b6ac46ac769f6f205419"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">TRTDataType</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">kernels</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">KVCacheIndex</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">UnderlyingType</span></span><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">value</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeIN7kernels12KVCacheIndexEE5valueE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeIN7kernels13FinishedStateEEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime11TRTDataTypeIN7kernels13FinishedStateEEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime11TRTDataTypeIN7kernels13FinishedStateEEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01kernels_1_1FinishedState_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TRTDataType</span></span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">kernels</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">FinishedState</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeIN7kernels13FinishedStateEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11TRTDataTypeIN7kernels13FinishedStateEE5valueE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11TRTDataTypeIN7kernels13FinishedStateEE5valueE"></span><span id="_CPPv2N12tensorrt_llm7runtime11TRTDataTypeIN7kernels13FinishedStateEE5valueE"></span><span id="tensorrt_llm::runtime::TRTDataType:kernels::FinishedState:::value__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01kernels_1_1FinishedState_01_4_1af0e2893266f9a1af9a97644162b45afa"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">TRTDataType</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">kernels</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">FinishedState</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">UnderlyingType</span></span><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">value</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeIN7kernels13FinishedStateEE5valueE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeIN7runtime11RequestTypeEEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime11TRTDataTypeIN7runtime11RequestTypeEEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime11TRTDataTypeIN7runtime11RequestTypeEEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01runtime_1_1RequestType_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TRTDataType</span></span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11RequestTypeE" title="tensorrt_llm::runtime::RequestType"><span class="n"><span class="pre">RequestType</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeIN7runtime11RequestTypeEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11TRTDataTypeIN7runtime11RequestTypeEE5valueE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11TRTDataTypeIN7runtime11RequestTypeEE5valueE"></span><span id="_CPPv2N12tensorrt_llm7runtime11TRTDataTypeIN7runtime11RequestTypeEE5valueE"></span><span id="tensorrt_llm::runtime::TRTDataType:runtime::RequestType:::value__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01runtime_1_1RequestType_01_4_1a689d61d98e3959c3f520274718c23541"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">TRTDataType</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">underlying_type_t</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11RequestTypeE" title="tensorrt_llm::runtime::RequestType"><span class="n"><span class="pre">RequestType</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">value</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeIN7runtime11RequestTypeEE5valueE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-typedefs">Typedefs</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime18PointerElementTypeE">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime18PointerElementTypeE"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime18PointerElementTypeE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="iBuffer_8h_1a6b78d5482ec51a801a9fe54db6eaa0f9"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">PointerElementType</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">remove_reference_t</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime18PointerElementTypeE" title="tensorrt_llm::runtime::PointerElementType::T"><span class="n"><span class="pre">T</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">element_type</span></span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime18PointerElementTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-enums">Enums</p>
|
||
<dl class="cpp enum-class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10MemoryTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10MemoryTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime10MemoryTypeE"></span><span class="target" id="iBuffer_8h_1a57051a59520fc335a72e6fd1d08fa869"></span><span class="k"><span class="pre">enum</span></span><span class="w"> </span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">MemoryType</span></span></span><span class="w"> </span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">int32_t</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p><em>Values:</em></p>
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10MemoryType4kGPUE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10MemoryType4kGPUE"></span><span id="_CPPv2N12tensorrt_llm7runtime10MemoryType4kGPUE"></span><span class="target" id="iBuffer_8h_1a57051a59520fc335a72e6fd1d08fa869a37ae524b76587efa776affdc5cdf2ac1"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kGPU</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10MemoryType4kGPUE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10MemoryType4kCPUE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10MemoryType4kCPUE"></span><span id="_CPPv2N12tensorrt_llm7runtime10MemoryType4kCPUE"></span><span class="target" id="iBuffer_8h_1a57051a59520fc335a72e6fd1d08fa869a2ce310327f474afc9f6774faa2f57903"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kCPU</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10MemoryType4kCPUE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10MemoryType7kPINNEDE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10MemoryType7kPINNEDE"></span><span id="_CPPv2N12tensorrt_llm7runtime10MemoryType7kPINNEDE"></span><span class="target" id="iBuffer_8h_1a57051a59520fc335a72e6fd1d08fa869ae3278bcaa387e6baeef9b80c1e61c35a"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kPINNED</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10MemoryType7kPINNEDE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10MemoryType4kUVME">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10MemoryType4kUVME"></span><span id="_CPPv2N12tensorrt_llm7runtime10MemoryType4kUVME"></span><span class="target" id="iBuffer_8h_1a57051a59520fc335a72e6fd1d08fa869a7d7de0bd70e1276255098b25010bdeb6"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kUVM</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10MemoryType4kUVME" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10MemoryType11kPINNEDPOOLE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10MemoryType11kPINNEDPOOLE"></span><span id="_CPPv2N12tensorrt_llm7runtime10MemoryType11kPINNEDPOOLE"></span><span class="target" id="iBuffer_8h_1a57051a59520fc335a72e6fd1d08fa869ac61fd7eec16ee67dfabffa6b6c7dd8aa"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kPINNEDPOOL</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10MemoryType11kPINNEDPOOLE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-functions">Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime16constPointerCastENSt10shared_ptrINSt14remove_const_tI1TEEEERKNSt10shared_ptrI1TEE">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime16constPointerCastERKNSt10shared_ptrI1TEE"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime16constPointerCastERKNSt10shared_ptrI1TEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="iBuffer_8h_1ae4deb6e27fee100e03742f3559ec25e7"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">remove_const_t</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime16constPointerCastENSt10shared_ptrINSt14remove_const_tI1TEEEERKNSt10shared_ptrI1TEE" title="tensorrt_llm::runtime::constPointerCast::T"><span class="n"><span class="pre">T</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">constPointerCast</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime16constPointerCastENSt10shared_ptrINSt14remove_const_tI1TEEEERKNSt10shared_ptrI1TEE" title="tensorrt_llm::runtime::constPointerCast::T"><span class="n"><span class="pre">T</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">ptr</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime16constPointerCastENSt10shared_ptrINSt14remove_const_tI1TEEEERKNSt10shared_ptrI1TEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I00EN12tensorrt_llm7runtime16constPointerCastENSt10shared_ptrINSt14remove_const_tI1TEEEERRNSt10unique_ptrI1T1DEE">
|
||
<span id="_CPPv3I00EN12tensorrt_llm7runtime16constPointerCastERRNSt10unique_ptrI1T1DEE"></span><span id="_CPPv2I00EN12tensorrt_llm7runtime16constPointerCastERRNSt10unique_ptrI1T1DEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">D</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="iBuffer_8h_1a1ea3ede13d88a2910f88551b802932b8"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">remove_const_t</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I00EN12tensorrt_llm7runtime16constPointerCastENSt10shared_ptrINSt14remove_const_tI1TEEEERRNSt10unique_ptrI1T1DEE" title="tensorrt_llm::runtime::constPointerCast::T"><span class="n"><span class="pre">T</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">constPointerCast</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">unique_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I00EN12tensorrt_llm7runtime16constPointerCastENSt10shared_ptrINSt14remove_const_tI1TEEEERRNSt10unique_ptrI1T1DEE" title="tensorrt_llm::runtime::constPointerCast::T"><span class="n"><span class="pre">T</span></span></a><span class="p"><span class="pre">,</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4I00EN12tensorrt_llm7runtime16constPointerCastENSt10shared_ptrINSt14remove_const_tI1TEEEERRNSt10unique_ptrI1T1DEE" title="tensorrt_llm::runtime::constPointerCast::D"><span class="n"><span class="pre">D</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">ptr</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4I00EN12tensorrt_llm7runtime16constPointerCastENSt10shared_ptrINSt14remove_const_tI1TEEEERRNSt10unique_ptrI1T1DEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime10bufferCastEPK1TRK7IBuffer">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime10bufferCastERK7IBuffer"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime10bufferCastERK7IBuffer"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="iBuffer_8h_1a582318a59a26e5e20f230a7d7c6a0d9f"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime10bufferCastEPK1TRK7IBuffer" title="tensorrt_llm::runtime::bufferCast::T"><span class="n"><span class="pre">T</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="sig-name descname"><span class="n"><span class="pre">bufferCast</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">buffer</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime10bufferCastEPK1TRK7IBuffer" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Gets a typed pointer to the constant underlying data of the buffer. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Template Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>T</strong> – The type of the underlying data. </p>
|
||
</dd>
|
||
<dt class="field-even">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p><strong>buffer</strong> – The buffer to get a pointer to. </p>
|
||
</dd>
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>A pointer to constant <code class="docutils literal notranslate"><span class="pre">T</span></code>. </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime10bufferCastEP1TR7IBuffer">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime10bufferCastER7IBuffer"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime10bufferCastER7IBuffer"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="iBuffer_8h_1a650f806d6eabc04b0cad8a5cdc888f4d"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime10bufferCastEP1TR7IBuffer" title="tensorrt_llm::runtime::bufferCast::T"><span class="n"><span class="pre">T</span></span></a><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="sig-name descname"><span class="n"><span class="pre">bufferCast</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">buffer</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime10bufferCastEP1TR7IBuffer" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Gets a typed pointer to the underlying data of the buffer. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Template Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>T</strong> – The type of the underlying data. </p>
|
||
</dd>
|
||
<dt class="field-even">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p><strong>buffer</strong> – The buffer to get a pointer to. </p>
|
||
</dd>
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>A pointer to <code class="docutils literal notranslate"><span class="pre">T</span></code>. </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEP1TRKN7IBuffer9SharedPtrE">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime16bufferCastOrNullERKN7IBuffer9SharedPtrE"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime16bufferCastOrNullERKN7IBuffer9SharedPtrE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="iBuffer_8h_1a3d2548426cec96db1d96d3dc2485431c"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEP1TRKN7IBuffer9SharedPtrE" title="tensorrt_llm::runtime::bufferCastOrNull::T"><span class="n"><span class="pre">T</span></span></a><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="sig-name descname"><span class="n"><span class="pre">bufferCastOrNull</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer9SharedPtrE" title="tensorrt_llm::runtime::IBuffer::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">bufferPtr</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEP1TRKN7IBuffer9SharedPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Retrieves a T typed pointer to the underlying data of the buffer pointed to by the bufferPtr, or nullptr if the bufferPtr is null. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Template Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>T</strong> – The type of the underlying data. </p>
|
||
</dd>
|
||
<dt class="field-even">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p><strong>bufferPtr</strong> – A possibly null shared ptr. </p>
|
||
</dd>
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>A pointer to T, possibly nullptr. </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEPK1TRKN7IBuffer14SharedConstPtrE">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime16bufferCastOrNullERKN7IBuffer14SharedConstPtrE"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime16bufferCastOrNullERKN7IBuffer14SharedConstPtrE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="iBuffer_8h_1ae558ab980f04f24da40bde948f7e27b2"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEPK1TRKN7IBuffer14SharedConstPtrE" title="tensorrt_llm::runtime::bufferCastOrNull::T"><span class="n"><span class="pre">T</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="sig-name descname"><span class="n"><span class="pre">bufferCastOrNull</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer14SharedConstPtrE" title="tensorrt_llm::runtime::IBuffer::SharedConstPtr"><span class="n"><span class="pre">SharedConstPtr</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">bufferPtr</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEPK1TRKN7IBuffer14SharedConstPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Retrieves a T const typed pointer to the underlying data of the buffer pointed to by the bufferPtr, or nullptr if the bufferPtr is null. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Template Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>T</strong> – The type of the underlying data. </p>
|
||
</dd>
|
||
<dt class="field-even">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p><strong>bufferPtr</strong> – A possibly null shared ptr. </p>
|
||
</dd>
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>A pointer to const T, possibly nullptr. </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEP1TRKNSt8optionalIN7IBuffer9SharedPtrEEE">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime16bufferCastOrNullERKNSt8optionalIN7IBuffer9SharedPtrEEE"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime16bufferCastOrNullERKNSt8optionalIN7IBuffer9SharedPtrEEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="iBuffer_8h_1a12d75e707c282d3abdd69933038489d6"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEP1TRKNSt8optionalIN7IBuffer9SharedPtrEEE" title="tensorrt_llm::runtime::bufferCastOrNull::T"><span class="n"><span class="pre">T</span></span></a><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="sig-name descname"><span class="n"><span class="pre">bufferCastOrNull</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer9SharedPtrE" title="tensorrt_llm::runtime::IBuffer::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">optionalBufferPtr</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEP1TRKNSt8optionalIN7IBuffer9SharedPtrEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Retrieves a T typed pointer to the underlying data of the buffer pointed to by the buffer pointer contained in the optionalBufferPtr, or nullptr if the optional doesn’t have a value. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Template Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>T</strong> – The type of the underlying data. </p>
|
||
</dd>
|
||
<dt class="field-even">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p><strong>optionalBufferPtr</strong> – A possibly empty optional. </p>
|
||
</dd>
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>A pointer to T, possibly nullptr. </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEPK1TRKNSt8optionalIN7IBuffer14SharedConstPtrEEE">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime16bufferCastOrNullERKNSt8optionalIN7IBuffer14SharedConstPtrEEE"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime16bufferCastOrNullERKNSt8optionalIN7IBuffer14SharedConstPtrEEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="iBuffer_8h_1aa1001e409507709767368842e3c6b0c7"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEPK1TRKNSt8optionalIN7IBuffer14SharedConstPtrEEE" title="tensorrt_llm::runtime::bufferCastOrNull::T"><span class="n"><span class="pre">T</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="sig-name descname"><span class="n"><span class="pre">bufferCastOrNull</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer14SharedConstPtrE" title="tensorrt_llm::runtime::IBuffer::SharedConstPtr"><span class="n"><span class="pre">SharedConstPtr</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">optionalBufferPtr</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime16bufferCastOrNullEPK1TRKNSt8optionalIN7IBuffer14SharedConstPtrEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Retrieves a T const typed pointer to the underlying data of the buffer pointed to by the buffer pointer contained in the optionalBufferPtr, or nullptr if the optional doesn’t have a value. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Template Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>T</strong> – The type of the underlying data. </p>
|
||
</dd>
|
||
<dt class="field-even">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p><strong>optionalBufferPtr</strong> – A possibly empty optional. </p>
|
||
</dd>
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>A pointer to const T, possibly nullptr. </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtimelsERNSt7ostreamERK7IBuffer">
|
||
<span id="_CPPv3N12tensorrt_llm7runtimelsERNSt7ostreamERK7IBuffer"></span><span id="_CPPv2N12tensorrt_llm7runtimelsERNSt7ostreamERK7IBuffer"></span><span id="tensorrt_llm::runtime::lshift-operator__osR.IBufferCR"></span><span class="target" id="iBuffer_8h_1a56e1be5ed31af23ee77c9001f938c1d7"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ostream</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="k"><span class="pre">operator</span></span><span class="o"><span class="pre"><<</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ostream</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">output</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">buffer</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtimelsERNSt7ostreamERK7IBuffer" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Utility function to print a buffer. </p>
|
||
</dd></dl>
|
||
|
||
</div>
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14BufferDataTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14BufferDataTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14BufferDataTypeE"></span><span id="tensorrt_llm::runtime::BufferDataType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferDataType"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">BufferDataType</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14BufferDataTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="docutils container">
|
||
<em>#include <iBuffer.h></em></div>
|
||
<p>A wrapper around <code class="docutils literal notranslate"><span class="pre">nvinfer1::DataType</span></code> that provides a support for pointer types. </p>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14BufferDataType14BufferDataTypeEN8nvinfer18DataTypeEbb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14BufferDataType14BufferDataTypeEN8nvinfer18DataTypeEbb"></span><span id="_CPPv2N12tensorrt_llm7runtime14BufferDataType14BufferDataTypeEN8nvinfer18DataTypeEbb"></span><span id="tensorrt_llm::runtime::BufferDataType::BufferDataType__nvinfer1::DataType.b.bCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferDataType_1a3cdb7e5d96b677f670d211cd3f0d2029"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">BufferDataType</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">dataType</span></span>, <span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">_unsigned</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">false</span></span>, <span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">pointer</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">false</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14BufferDataType14BufferDataTypeEN8nvinfer18DataTypeEbb" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14BufferDataTypecvN8nvinfer18DataTypeEEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14BufferDataTypecvN8nvinfer18DataTypeEEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14BufferDataTypecvN8nvinfer18DataTypeEEv"></span><span id="tensorrt_llm::runtime::BufferDataType::castto-nvinfer1::DataType-operatorCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferDataType_1ae0e234f2e49d57ea1ec39bf76ef5f843"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="k"><span class="pre">operator</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime14BufferDataTypecvN8nvinfer18DataTypeEEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14BufferDataType11getDataTypeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14BufferDataType11getDataTypeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14BufferDataType11getDataTypeEv"></span><span id="tensorrt_llm::runtime::BufferDataType::getDataTypeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferDataType_1a375ef9858082b42a89a8fbddeb758836"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getDataType</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime14BufferDataType11getDataTypeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14BufferDataType9isPointerEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14BufferDataType9isPointerEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14BufferDataType9isPointerEv"></span><span id="tensorrt_llm::runtime::BufferDataType::isPointerCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferDataType_1a4f456474dbc06fcfb4c5fdd4f30c681b"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">isPointer</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime14BufferDataType9isPointerEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14BufferDataType10isUnsignedEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14BufferDataType10isUnsignedEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14BufferDataType10isUnsignedEv"></span><span id="tensorrt_llm::runtime::BufferDataType::isUnsignedCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferDataType_1a84da512a92aa9b9ba7498d8eface36d1"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">isUnsigned</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime14BufferDataType10isUnsignedEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14BufferDataType7getSizeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14BufferDataType7getSizeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14BufferDataType7getSizeEv"></span><span id="tensorrt_llm::runtime::BufferDataType::getSizeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferDataType_1a0e1c8d7eb345611735b5c8998efc1b29"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getSize</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime14BufferDataType7getSizeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14BufferDataType13getSizeInBitsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14BufferDataType13getSizeInBitsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14BufferDataType13getSizeInBitsEv"></span><span id="tensorrt_llm::runtime::BufferDataType::getSizeInBitsCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferDataType_1a42aaae0de1ad95e54048fa773c1cf1e8"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getSizeInBits</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime14BufferDataType13getSizeInBitsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14BufferDataType15kTrtPointerTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14BufferDataType15kTrtPointerTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14BufferDataType15kTrtPointerTypeE"></span><span id="tensorrt_llm::runtime::BufferDataType::kTrtPointerType__auto"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferDataType_1ac186826cc80422eae28ba74309100cd9"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kTrtPointerType</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">kINT64</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14BufferDataType15kTrtPointerTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-members">Private Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14BufferDataType9mDataTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14BufferDataType9mDataTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14BufferDataType9mDataTypeE"></span><span id="tensorrt_llm::runtime::BufferDataType::mDataType__nvinfer1::DataType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferDataType_1a57916add2426171d7066d14df4e796d8"></span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mDataType</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14BufferDataType9mDataTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14BufferDataType9mUnsignedE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14BufferDataType9mUnsignedE"></span><span id="_CPPv2N12tensorrt_llm7runtime14BufferDataType9mUnsignedE"></span><span id="tensorrt_llm::runtime::BufferDataType::mUnsigned__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferDataType_1a675fe3124bb2de4dd047ce69a1a5a88b"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mUnsigned</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14BufferDataType9mUnsignedE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14BufferDataType8mPointerE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14BufferDataType8mPointerE"></span><span id="_CPPv2N12tensorrt_llm7runtime14BufferDataType8mPointerE"></span><span id="tensorrt_llm::runtime::BufferDataType::mPointer__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferDataType_1a494f3b4c8e08ef43209cac0e2f114e0c"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mPointer</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14BufferDataType8mPointerE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime11BufferRangeE">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime11BufferRangeE"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime11BufferRangeE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferRange"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">BufferRange</span></span></span><span class="w"> </span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="k"><span class="pre">public</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">common</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ArrayView</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime11BufferRangeE" title="tensorrt_llm::runtime::BufferRange::T"><span class="n"><span class="pre">T</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime11BufferRangeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11BufferRange4BaseE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11BufferRange4BaseE"></span><span id="_CPPv2N12tensorrt_llm7runtime11BufferRange4BaseE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferRange_1a58ffed28eecc57b7ad0178ce58cdaa61"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Base</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">common</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ArrayView</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime11BufferRangeE" title="tensorrt_llm::runtime::BufferRange::T"><span class="n"><span class="pre">T</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange4BaseE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11BufferRange11BufferRangeEP1T9size_type">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11BufferRange11BufferRangeEP1T9size_type"></span><span id="_CPPv2N12tensorrt_llm7runtime11BufferRange11BufferRangeEP1T9size_type"></span><span id="tensorrt_llm::runtime::BufferRange::BufferRange__TP.size_type"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferRange_1a7ea39c3ef4a0a2bf24228697045fc145"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">BufferRange</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime11BufferRangeE" title="tensorrt_llm::runtime::BufferRange::T"><span class="n"><span class="pre">T</span></span></a><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="n sig-param"><span class="pre">data</span></span>, <span class="n"><span class="pre">size_type</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange11BufferRangeEP1T9size_type" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0_NSt11enable_if_tIXntNSt10is_const_vI1UEEEbEEEN12tensorrt_llm7runtime11BufferRange11BufferRangeER7IBuffer">
|
||
<span id="_CPPv3I0_NSt11enable_if_tIXntNSt10is_const_vI1UEEEbEEEN12tensorrt_llm7runtime11BufferRange11BufferRangeER7IBuffer"></span><span id="_CPPv2I0_NSt11enable_if_tIX!std::is_const_v<U>EbEEEN12tensorrt_llm7runtime11BufferRange11BufferRangeER7IBuffer"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">U</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime11BufferRangeE" title="tensorrt_llm::runtime::BufferRange::T"><span class="n"><span class="pre">T</span></span></a><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">enable_if_t</span></span><span class="p"><span class="pre"><</span></span><span class="o"><span class="pre">!</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">is_const_v</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tIXntNSt10is_const_vI1UEEEbEEEN12tensorrt_llm7runtime11BufferRange11BufferRangeER7IBuffer" title="tensorrt_llm::runtime::BufferRange::BufferRange::U"><span class="n"><span class="pre">U</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">true</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferRange_1aa31b97783316f3d1e92432c0753c9e99"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">BufferRange</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">buffer</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0_NSt11enable_if_tIXntNSt10is_const_vI1UEEEbEEEN12tensorrt_llm7runtime11BufferRange11BufferRangeER7IBuffer" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0_NSt11enable_if_tINSt10is_const_vI1UEEbEEEN12tensorrt_llm7runtime11BufferRange11BufferRangeERK7IBuffer">
|
||
<span id="_CPPv3I0_NSt11enable_if_tINSt10is_const_vI1UEEbEEEN12tensorrt_llm7runtime11BufferRange11BufferRangeERK7IBuffer"></span><span id="_CPPv2I0_NSt11enable_if_tINSt10is_const_vI1UEEbEEEN12tensorrt_llm7runtime11BufferRange11BufferRangeERK7IBuffer"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">U</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime11BufferRangeE" title="tensorrt_llm::runtime::BufferRange::T"><span class="n"><span class="pre">T</span></span></a><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">enable_if_t</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">is_const_v</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI1UEEbEEEN12tensorrt_llm7runtime11BufferRange11BufferRangeERK7IBuffer" title="tensorrt_llm::runtime::BufferRange::BufferRange::U"><span class="n"><span class="pre">U</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">true</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferRange_1a5bb8dc187700342fb7254683285de05a"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">BufferRange</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">buffer</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI1UEEbEEEN12tensorrt_llm7runtime11BufferRange11BufferRangeERK7IBuffer" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I_N8nvinfer18DataTypeE_b_bEN12tensorrt_llm7runtime14DataTypeTraitsE">
|
||
<span id="_CPPv3I_N8nvinfer18DataTypeE_b_bEN12tensorrt_llm7runtime14DataTypeTraitsE"></span><span id="_CPPv2I_N8nvinfer18DataTypeE_b_bEN12tensorrt_llm7runtime14DataTypeTraitsE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kDataType</span></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kIsUnsigned</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">false</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kIsPointer</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">false</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">DataTypeTraits</span></span></span><a class="headerlink" href="#_CPPv4I_N8nvinfer18DataTypeE_b_bEN12tensorrt_llm7runtime14DataTypeTraitsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="docutils container">
|
||
<em>#include <iBuffer.h></em></div>
|
||
<p>For converting a TensorRT data type to a C++ data type. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I_N8nvinfer18DataTypeE_bEN12tensorrt_llm7runtime14DataTypeTraitsI9kDataType9kUnsignedXL1EEEE">
|
||
<span id="_CPPv3I_N8nvinfer18DataTypeE_bEN12tensorrt_llm7runtime14DataTypeTraitsI9kDataType9kUnsignedXL1EEEE"></span><span id="_CPPv2I_N8nvinfer18DataTypeE_bEN12tensorrt_llm7runtime14DataTypeTraitsI9kDataType9kUnsignedXtrueEEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kDataType</span></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kUnsigned</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01kDataType_00_01kUnsigned_00_01true_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">DataTypeTraits</span></span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I_N8nvinfer18DataTypeE_bEN12tensorrt_llm7runtime14DataTypeTraitsI9kDataType9kUnsignedXL1EEEE" title="tensorrt_llm::runtime::DataTypeTraits<kDataType, kUnsigned, true>::kDataType"><span class="n"><span class="pre">kDataType</span></span></a><span class="p"><span class="pre">,</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4I_N8nvinfer18DataTypeE_bEN12tensorrt_llm7runtime14DataTypeTraitsI9kDataType9kUnsignedXL1EEEE" title="tensorrt_llm::runtime::DataTypeTraits<kDataType, kUnsigned, true>::kUnsigned"><span class="n"><span class="pre">kUnsigned</span></span></a><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="k"><span class="pre">true</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4I_N8nvinfer18DataTypeE_bEN12tensorrt_llm7runtime14DataTypeTraitsI9kDataType9kUnsignedXL1EEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsI9kDataType9kUnsignedXL1EEE4typeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsI9kDataType9kUnsignedXL1EEE4typeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsI9kDataType9kUnsignedXtrueEE4typeE"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01kDataType_00_01kUnsigned_00_01true_01_4_1a4e4f2844614eff1320acd710bc0cfe9c"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">type</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="n"><span class="pre">DataTypeTraits</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I_N8nvinfer18DataTypeE_bEN12tensorrt_llm7runtime14DataTypeTraitsI9kDataType9kUnsignedXL1EEEE" title="tensorrt_llm::runtime::DataTypeTraits<kDataType, kUnsigned, true>::kDataType"><span class="n"><span class="pre">kDataType</span></span></a><span class="p"><span class="pre">,</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4I_N8nvinfer18DataTypeE_bEN12tensorrt_llm7runtime14DataTypeTraitsI9kDataType9kUnsignedXL1EEEE" title="tensorrt_llm::runtime::DataTypeTraits<kDataType, kUnsigned, true>::kUnsigned"><span class="n"><span class="pre">kUnsigned</span></span></a><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="k"><span class="pre">false</span></span><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">type</span></span><span class="p"><span class="pre">*</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsI9kDataType9kUnsignedXL1EEE4typeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsI9kDataType9kUnsignedXL1EEE4nameE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsI9kDataType9kUnsignedXL1EEE4nameE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsI9kDataType9kUnsignedXtrueEE4nameE"></span><span id="tensorrt_llm::runtime::DataTypeTraits:kDataType.kUnsigned.true:::name__cA"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01kDataType_00_01kUnsigned_00_01true_01_4_1a3a25a81fd2f6753350919c8a0be4f039"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">char</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">name</span></span></span><span class="p"><span class="pre">[</span></span><span class="p"><span class="pre">]</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"*"</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsI9kDataType9kUnsignedXL1EEE4nameE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsI9kDataType9kUnsignedXL1EEE4sizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DataTypeTraitsI9kDataType9kUnsignedXL1EEE4sizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DataTypeTraitsI9kDataType9kUnsignedXtrueEE4sizeE"></span><span id="tensorrt_llm::runtime::DataTypeTraits:kDataType.kUnsigned.true:::size__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01kDataType_00_01kUnsigned_00_01true_01_4_1a1d99943821b4bdd36c1aec9f911ee992"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">size</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">sizeof</span></span><span class="p"><span class="pre">(</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsI9kDataType9kUnsignedXL1EEE4typeE" title="tensorrt_llm::runtime::DataTypeTraits<kDataType, kUnsigned, true>::type"><span class="n"><span class="pre">type</span></span></a><span class="p"><span class="pre">)</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DataTypeTraitsI9kDataType9kUnsignedXL1EEE4sizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kBOOL_00_01kUnsigned_01_4"></span><span class="sig-name descname"><span class="pre">template<bool</span> <span class="pre">kUnsigned></span> <span class="pre">kBOOL,</span> <span class="pre">kUnsigned</span> <span class="pre">></span></span></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime19PhonyNameDueToError4typeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime19PhonyNameDueToError4typeE"></span><span id="_CPPv2N12tensorrt_llm7runtime19PhonyNameDueToError4typeE"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kBOOL_00_01kUnsigned_01_4_1a3ef7ec1c22f6cc0b397ab5eb28c232eb"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">type</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime19PhonyNameDueToError4typeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime19PhonyNameDueToError4nameE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime19PhonyNameDueToError4nameE"></span><span id="_CPPv2N12tensorrt_llm7runtime19PhonyNameDueToError4nameE"></span><span id="tensorrt_llm::runtime::PhonyNameDueToError::name__cA"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kBOOL_00_01kUnsigned_01_4_1a60ce7e6d8364224764d2415e4b709925"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">char</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">name</span></span></span><span class="p"><span class="pre">[</span></span><span class="p"><span class="pre">]</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"bool"</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime19PhonyNameDueToError4nameE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime19PhonyNameDueToError4sizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime19PhonyNameDueToError4sizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime19PhonyNameDueToError4sizeE"></span><span id="tensorrt_llm::runtime::PhonyNameDueToError::size__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kBOOL_00_01kUnsigned_01_4_1aea0d24ee412e727b1e7d1831501600df"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">size</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">sizeof</span></span><span class="p"><span class="pre">(</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19PhonyNameDueToError4typeE" title="tensorrt_llm::runtime::PhonyNameDueToError::type"><span class="n"><span class="pre">type</span></span></a><span class="p"><span class="pre">)</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime19PhonyNameDueToError4sizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kFLOAT_01_4"></span><span class="sig-name descname"><span class="pre">template<></span> <span class="pre">kFLOAT</span> <span class="pre">></span></span></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kFLOAT_01_4_1af6ef0c2bf073670006e1a68fd861495b"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">type</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="kt"><span class="pre">float</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kFLOAT_01_4_1a7253109d1e4c01ed33157460dba35ec7"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">char</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">name</span></span></span><span class="p"><span class="pre">[</span></span><span class="p"><span class="pre">]</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"float"</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kFLOAT_01_4_1afff1a8fde1dc21c665c8e1f3f8cc3fbd"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">size</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">sizeof</span></span><span class="p"><span class="pre">(</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19PhonyNameDueToError4typeE" title="tensorrt_llm::runtime::PhonyNameDueToError::type"><span class="n"><span class="pre">type</span></span></a><span class="p"><span class="pre">)</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kHALF_01_4"></span><span class="sig-name descname"><span class="pre">template<></span> <span class="pre">kHALF</span> <span class="pre">></span></span></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kHALF_01_4_1a08c9427d16066ff8ee9e9deb35a91871"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">type</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">half</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kHALF_01_4_1a661db84e273578749fbd43d1f6829aa0"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">char</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">name</span></span></span><span class="p"><span class="pre">[</span></span><span class="p"><span class="pre">]</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"half"</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kHALF_01_4_1aad4af9eb10a14553b0dc46bd210454c1"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">size</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">sizeof</span></span><span class="p"><span class="pre">(</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19PhonyNameDueToError4typeE" title="tensorrt_llm::runtime::PhonyNameDueToError::type"><span class="n"><span class="pre">type</span></span></a><span class="p"><span class="pre">)</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT32_01_4"></span><span class="sig-name descname"><span class="pre">template<></span> <span class="pre">kINT32</span> <span class="pre">></span></span></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT32_01_4_1a37fa1b65a6bece3eb638f7693132053c"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">type</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">int32_t</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT32_01_4_1a98f4973f56a92c96b5bd1c4298b4ca72"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">char</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">name</span></span></span><span class="p"><span class="pre">[</span></span><span class="p"><span class="pre">]</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"int32"</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT32_01_4_1ad02b1cf34aed9754967e7bb76a55fb0b"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">size</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">sizeof</span></span><span class="p"><span class="pre">(</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19PhonyNameDueToError4typeE" title="tensorrt_llm::runtime::PhonyNameDueToError::type"><span class="n"><span class="pre">type</span></span></a><span class="p"><span class="pre">)</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT32_00_01true_01_4"></span><span class="sig-name descname"><span class="pre">template<></span> <span class="pre">kINT32,</span> <span class="pre">true</span> <span class="pre">></span></span></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT32_00_01true_01_4_1a41a0f3bcdb82b843a9594fbe518c148d"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">type</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">uint32_t</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT32_00_01true_01_4_1a455a075a88a89b0bf29b5ac1afc54320"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">char</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">name</span></span></span><span class="p"><span class="pre">[</span></span><span class="p"><span class="pre">]</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"uint32"</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT32_00_01true_01_4_1a420375ac7c713ee26156c7e143327393"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">size</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">sizeof</span></span><span class="p"><span class="pre">(</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19PhonyNameDueToError4typeE" title="tensorrt_llm::runtime::PhonyNameDueToError::type"><span class="n"><span class="pre">type</span></span></a><span class="p"><span class="pre">)</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT64_01_4"></span><span class="sig-name descname"><span class="pre">template<></span> <span class="pre">kINT64</span> <span class="pre">></span></span></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT64_01_4_1a288b116cbff3532cf2b80e7206277846"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">type</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">int64_t</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT64_01_4_1a912e86b1f6bb4572292b27ed87819384"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">char</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">name</span></span></span><span class="p"><span class="pre">[</span></span><span class="p"><span class="pre">]</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"int64"</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT64_01_4_1ac9994f6271fd3dd86e3be01df67a5577"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">size</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">sizeof</span></span><span class="p"><span class="pre">(</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19PhonyNameDueToError4typeE" title="tensorrt_llm::runtime::PhonyNameDueToError::type"><span class="n"><span class="pre">type</span></span></a><span class="p"><span class="pre">)</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT64_00_01true_01_4"></span><span class="sig-name descname"><span class="pre">template<></span> <span class="pre">kINT64,</span> <span class="pre">true</span> <span class="pre">></span></span></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT64_00_01true_01_4_1a8376cfb6029bfec1e7aaa9c29369b094"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">type</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">uint64_t</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT64_00_01true_01_4_1ae082c5ad1c2c409b8cbedc64b69e7f1d"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">char</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">name</span></span></span><span class="p"><span class="pre">[</span></span><span class="p"><span class="pre">]</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"uint64"</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT64_00_01true_01_4_1af5140fb67b9b5cae4c40ea6f998f1223"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">size</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">sizeof</span></span><span class="p"><span class="pre">(</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19PhonyNameDueToError4typeE" title="tensorrt_llm::runtime::PhonyNameDueToError::type"><span class="n"><span class="pre">type</span></span></a><span class="p"><span class="pre">)</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT8_01_4"></span><span class="sig-name descname"><span class="pre">template<></span> <span class="pre">kINT8</span> <span class="pre">></span></span></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT8_01_4_1a51b8ee14d86b4b2a6cc5e57a1d524c43"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">type</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">int8_t</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT8_01_4_1ad45ffa244d70cff61abcc7c89436e839"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">char</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">name</span></span></span><span class="p"><span class="pre">[</span></span><span class="p"><span class="pre">]</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"int8"</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kINT8_01_4_1ad29e96b3fe577dba8524735a4f42f768"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">size</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">sizeof</span></span><span class="p"><span class="pre">(</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19PhonyNameDueToError4typeE" title="tensorrt_llm::runtime::PhonyNameDueToError::type"><span class="n"><span class="pre">type</span></span></a><span class="p"><span class="pre">)</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kUINT8_00_01kUnsigned_01_4"></span><span class="sig-name descname"><span class="pre">template<bool</span> <span class="pre">kUnsigned></span> <span class="pre">kUINT8,</span> <span class="pre">kUnsigned</span> <span class="pre">></span></span></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kUINT8_00_01kUnsigned_01_4_1a5a3089fdad68ca4ce8bf36137d18a147"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">type</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">uint8_t</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kUINT8_00_01kUnsigned_01_4_1a997c0219602be3dfcde787c95462e1d5"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">char</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">name</span></span></span><span class="p"><span class="pre">[</span></span><span class="p"><span class="pre">]</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"uint8"</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1DataTypeTraits_3_01nvinfer1_1_1DataType_1_1kUINT8_00_01kUnsigned_01_4_1abb3870f3b3387d0cfedb4af98f5d71a6"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">size</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">sizeof</span></span><span class="p"><span class="pre">(</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19PhonyNameDueToError4typeE" title="tensorrt_llm::runtime::PhonyNameDueToError::type"><span class="n"><span class="pre">type</span></span></a><span class="p"><span class="pre">)</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7IBufferE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7IBufferE"></span><span id="_CPPv2N12tensorrt_llm7runtime7IBufferE"></span><span id="tensorrt_llm::runtime::IBuffer"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">IBuffer</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Subclassed by <a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1ITensor"><span class="std std-ref">tensorrt_llm::runtime::ITensor</span></a></p>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7IBuffer9UniquePtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7IBuffer9UniquePtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime7IBuffer9UniquePtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1ab5684b52d1990c0a0f70c25d0b6c1465"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">UniquePtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">unique_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer9UniquePtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7IBuffer9SharedPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7IBuffer9SharedPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime7IBuffer9SharedPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1a2f0c3696d87a3887b38da126cbef1759"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">SharedPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer9SharedPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7IBuffer14UniqueConstPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7IBuffer14UniqueConstPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime7IBuffer14UniqueConstPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1a3892842dc7c9dcc1cad622470ac97999"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">UniqueConstPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">unique_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer14UniqueConstPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7IBuffer14SharedConstPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7IBuffer14SharedConstPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime7IBuffer14SharedConstPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1a8eefcf151b57310d8ca78b8e25c46c39"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">SharedConstPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer14SharedConstPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7IBuffer8DataTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7IBuffer8DataTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime7IBuffer8DataTypeE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1a4036935d2bbb442bf9d071df2eff24bd"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">DataType</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer8DataTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7IBuffer4dataEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7IBuffer4dataEv"></span><span id="_CPPv2N12tensorrt_llm7runtime7IBuffer4dataEv"></span><span id="tensorrt_llm::runtime::IBuffer::data"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1a56ee6dcc7c3b905e6b929e3e56041874"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="sig-name descname"><span class="n"><span class="pre">data</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer4dataEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Returns a pointer to underlying array. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime7IBuffer4dataEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime7IBuffer4dataEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime7IBuffer4dataEv"></span><span id="tensorrt_llm::runtime::IBuffer::dataC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1ab18c55dc858ab44df1216f9f1bce157c"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="sig-name descname"><span class="n"><span class="pre">data</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime7IBuffer4dataEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Returns a pointer to underlying array. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7IBuffer4dataENSt6size_tE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7IBuffer4dataENSt6size_tE"></span><span id="_CPPv2N12tensorrt_llm7runtime7IBuffer4dataENSt6size_tE"></span><span id="tensorrt_llm::runtime::IBuffer::data__std::s"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1a6f57afc5472f47ec5c76ce5cddb16849"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="sig-name descname"><span class="n"><span class="pre">data</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">index</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer4dataENSt6size_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Returns a pointer to the underlying array at a given element index. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime7IBuffer4dataENSt6size_tE">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime7IBuffer4dataENSt6size_tE"></span><span id="_CPPv2NK12tensorrt_llm7runtime7IBuffer4dataENSt6size_tE"></span><span id="tensorrt_llm::runtime::IBuffer::data__std::sC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1ab59d0db536d7cb31593cec33e588fc43"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="sig-name descname"><span class="n"><span class="pre">data</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">index</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime7IBuffer4dataENSt6size_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Returns a pointer to the underlying array at a given element index. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime7IBuffer7getSizeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime7IBuffer7getSizeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime7IBuffer7getSizeEv"></span><span id="tensorrt_llm::runtime::IBuffer::getSizeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1ab280c88977e5cf7d3c3ab88e29dcbe62"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getSize</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime7IBuffer7getSizeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Returns the size (in number of elements) of the buffer. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime7IBuffer14getSizeInBytesEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime7IBuffer14getSizeInBytesEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime7IBuffer14getSizeInBytesEv"></span><span id="tensorrt_llm::runtime::IBuffer::getSizeInBytesC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1a3121c0295ee572bb45067bba2d35defa"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getSizeInBytes</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime7IBuffer14getSizeInBytesEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Returns the size (in bytes) of the buffer. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime7IBuffer11getCapacityEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime7IBuffer11getCapacityEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime7IBuffer11getCapacityEv"></span><span id="tensorrt_llm::runtime::IBuffer::getCapacityC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1a33c48c142ee4ee1c1537d1bef5a7ed16"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getCapacity</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime7IBuffer11getCapacityEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Returns the capacity of the buffer. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime7IBuffer11getDataTypeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime7IBuffer11getDataTypeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime7IBuffer11getDataTypeEv"></span><span id="tensorrt_llm::runtime::IBuffer::getDataTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1a006dc512c6b37d582a2e825249c4a3a2"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer8DataTypeE" title="tensorrt_llm::runtime::IBuffer::DataType"><span class="n"><span class="pre">DataType</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getDataType</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime7IBuffer11getDataTypeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Returns the data type of the buffer. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime7IBuffer15getDataTypeNameEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime7IBuffer15getDataTypeNameEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime7IBuffer15getDataTypeNameEv"></span><span id="tensorrt_llm::runtime::IBuffer::getDataTypeNameC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1a8aa7843674fa7d71b1c0e894a312bd94"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">char</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="sig-name descname"><span class="n"><span class="pre">getDataTypeName</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime7IBuffer15getDataTypeNameEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime7IBuffer13getMemoryTypeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime7IBuffer13getMemoryTypeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime7IBuffer13getMemoryTypeEv"></span><span id="tensorrt_llm::runtime::IBuffer::getMemoryTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1ade5fc6e89a07ad03927a616fdcbb0463"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="tensorrt_llm::runtime::MemoryType"><span class="n"><span class="pre">MemoryType</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getMemoryType</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime7IBuffer13getMemoryTypeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Returns the memory type of the buffer. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime7IBuffer17getMemoryTypeNameEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime7IBuffer17getMemoryTypeNameEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime7IBuffer17getMemoryTypeNameEv"></span><span id="tensorrt_llm::runtime::IBuffer::getMemoryTypeNameC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1a9108d1411aea38779b0aa4234250c9f5"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">char</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="sig-name descname"><span class="n"><span class="pre">getMemoryTypeName</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime7IBuffer17getMemoryTypeNameEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7IBuffer6resizeENSt6size_tE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7IBuffer6resizeENSt6size_tE"></span><span id="_CPPv2N12tensorrt_llm7runtime7IBuffer6resizeENSt6size_tE"></span><span id="tensorrt_llm::runtime::IBuffer::resize__std::s"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1af1281049c93d0a1d61b8682170761ad4"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">resize</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">newSize</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer6resizeENSt6size_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Resizes the buffer. This is a no-op if the new size is smaller than or equal to the current capacity. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7IBuffer7releaseEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7IBuffer7releaseEv"></span><span id="_CPPv2N12tensorrt_llm7runtime7IBuffer7releaseEv"></span><span id="tensorrt_llm::runtime::IBuffer::release"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1a51dc4a186d9b315dfe77aacd33677ff7"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">release</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer7releaseEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Releases the buffer. It will be reset to nullptr. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7IBufferD0Ev">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7IBufferD0Ev"></span><span id="_CPPv2N12tensorrt_llm7runtime7IBufferD0Ev"></span><span id="tensorrt_llm::runtime::IBuffer::~IBuffer"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1abeb460077884adc9aec0c351c9ef3637"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">~IBuffer</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">default</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7IBufferD0Ev" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7IBuffer7IBufferERK7IBuffer">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7IBuffer7IBufferERK7IBuffer"></span><span id="_CPPv2N12tensorrt_llm7runtime7IBuffer7IBufferERK7IBuffer"></span><span id="tensorrt_llm::runtime::IBuffer::IBuffer__IBufferCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1ac9fd5dfcd8486d8f8150fdb470c38055"></span><span class="sig-name descname"><span class="n"><span class="pre">IBuffer</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer7IBufferERK7IBuffer" title="tensorrt_llm::runtime::IBuffer::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="p"><span class="pre">&</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">delete</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer7IBufferERK7IBuffer" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Not allowed to copy. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7IBufferaSERK7IBuffer">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7IBufferaSERK7IBuffer"></span><span id="_CPPv2N12tensorrt_llm7runtime7IBufferaSERK7IBuffer"></span><span id="tensorrt_llm::runtime::IBuffer::assign-operator__IBufferCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1ac2d5c67a705ab550d6aaecd01108408e"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="k"><span class="pre">operator</span></span><span class="o"><span class="pre">=</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBufferE" title="tensorrt_llm::runtime::IBuffer"><span class="n"><span class="pre">IBuffer</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="p"><span class="pre">&</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">delete</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7IBufferaSERK7IBuffer" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Not allowed to copy. </p>
|
||
</dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-functions">Public Static Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7IBuffer15getDataTypeNameE8DataType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7IBuffer15getDataTypeNameE8DataType"></span><span id="_CPPv2N12tensorrt_llm7runtime7IBuffer15getDataTypeNameE8DataType"></span><span id="tensorrt_llm::runtime::IBuffer::getDataTypeName__DataType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1afda51ea16de70b983603ca5e6225e255"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">char</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="sig-name descname"><span class="n"><span class="pre">getDataTypeName</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer8DataTypeE" title="tensorrt_llm::runtime::IBuffer::DataType"><span class="n"><span class="pre">DataType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">dataType</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer15getDataTypeNameE8DataType" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7IBuffer5sliceE9SharedPtrNSt6size_tENSt6size_tE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7IBuffer5sliceE9SharedPtrNSt6size_tENSt6size_tE"></span><span id="_CPPv2N12tensorrt_llm7runtime7IBuffer5sliceE9SharedPtrNSt6size_tENSt6size_tE"></span><span id="tensorrt_llm::runtime::IBuffer::slice__SharedPtr.std::s.std::s"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1a6b9aed7e66f2bfc1f3f710d12dbceea6"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer9UniquePtrE" title="tensorrt_llm::runtime::IBuffer::UniquePtr"><span class="n"><span class="pre">UniquePtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">slice</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer9SharedPtrE" title="tensorrt_llm::runtime::IBuffer::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">buffer</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">offset</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer5sliceE9SharedPtrNSt6size_tENSt6size_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Creates a sliced view on the underlying <code class="docutils literal notranslate"><span class="pre">buffer</span></code>. The view will have the same data type as <code class="docutils literal notranslate"><span class="pre">buffer</span></code>. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><ul class="simple">
|
||
<li><p><strong>buffer</strong> – The buffer to view. </p></li>
|
||
<li><p><strong>offset</strong> – The offset of the view. </p></li>
|
||
<li><p><strong>size</strong> – The size of the view. </p></li>
|
||
</ul>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>A view on the <code class="docutils literal notranslate"><span class="pre">buffer</span></code>. </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7IBuffer5sliceE14UniqueConstPtrRR9TConstPtrNSt6size_tENSt6size_tE">
|
||
<span id="_CPPv3I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7IBuffer5sliceERR9TConstPtrNSt6size_tENSt6size_tE"></span><span id="_CPPv2I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7IBuffer5sliceERR9TConstPtrNSt6size_tENSt6size_tE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TConstPtr</span></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">enable_if_t</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">is_const_v</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime18PointerElementTypeE" title="tensorrt_llm::runtime::PointerElementType"><span class="n"><span class="pre">PointerElementType</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7IBuffer5sliceE14UniqueConstPtrRR9TConstPtrNSt6size_tENSt6size_tE" title="tensorrt_llm::runtime::IBuffer::slice::TConstPtr"><span class="n"><span class="pre">TConstPtr</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="kt"><span class="pre">int</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1a40f5f57b98c9918b617c1de832390b9c"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer14UniqueConstPtrE" title="tensorrt_llm::runtime::IBuffer::UniqueConstPtr"><span class="n"><span class="pre">UniqueConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">slice</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7IBuffer5sliceE14UniqueConstPtrRR9TConstPtrNSt6size_tENSt6size_tE" title="tensorrt_llm::runtime::IBuffer::slice::TConstPtr"><span class="n"><span class="pre">TConstPtr</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">tensor</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">offset</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7IBuffer5sliceE14UniqueConstPtrRR9TConstPtrNSt6size_tENSt6size_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7IBuffer5sliceE9SharedPtrNSt6size_tE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7IBuffer5sliceE9SharedPtrNSt6size_tE"></span><span id="_CPPv2N12tensorrt_llm7runtime7IBuffer5sliceE9SharedPtrNSt6size_tE"></span><span id="tensorrt_llm::runtime::IBuffer::slice__SharedPtr.std::s"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1a310d420f890b10b16951eae4b0227d21"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer9UniquePtrE" title="tensorrt_llm::runtime::IBuffer::UniquePtr"><span class="n"><span class="pre">UniquePtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">slice</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer9SharedPtrE" title="tensorrt_llm::runtime::IBuffer::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">buffer</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">offset</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer5sliceE9SharedPtrNSt6size_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7IBuffer5sliceE14UniqueConstPtrRR9TConstPtrNSt6size_tE">
|
||
<span id="_CPPv3I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7IBuffer5sliceERR9TConstPtrNSt6size_tE"></span><span id="_CPPv2I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7IBuffer5sliceERR9TConstPtrNSt6size_tE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TConstPtr</span></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">enable_if_t</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">is_const_v</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime18PointerElementTypeE" title="tensorrt_llm::runtime::PointerElementType"><span class="n"><span class="pre">PointerElementType</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7IBuffer5sliceE14UniqueConstPtrRR9TConstPtrNSt6size_tE" title="tensorrt_llm::runtime::IBuffer::slice::TConstPtr"><span class="n"><span class="pre">TConstPtr</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="kt"><span class="pre">int</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1a9ffdc0f354482f983b6a16825ad25109"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer14UniqueConstPtrE" title="tensorrt_llm::runtime::IBuffer::UniqueConstPtr"><span class="n"><span class="pre">UniqueConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">slice</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7IBuffer5sliceE14UniqueConstPtrRR9TConstPtrNSt6size_tE" title="tensorrt_llm::runtime::IBuffer::slice::TConstPtr"><span class="n"><span class="pre">TConstPtr</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">tensor</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">offset</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7IBuffer5sliceE14UniqueConstPtrRR9TConstPtrNSt6size_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7IBuffer4viewE9SharedPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7IBuffer4viewE9SharedPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime7IBuffer4viewE9SharedPtr"></span><span id="tensorrt_llm::runtime::IBuffer::view__SharedPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1a4fb6a8d4a92376cdea0957a26629f53b"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer9UniquePtrE" title="tensorrt_llm::runtime::IBuffer::UniquePtr"><span class="n"><span class="pre">UniquePtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">view</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer9SharedPtrE" title="tensorrt_llm::runtime::IBuffer::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tensor</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer4viewE9SharedPtr" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Returns a view on the underlying <code class="docutils literal notranslate"><span class="pre">tensor</span></code> which can be independently resized. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>tensor</strong> – The tensor to view. </p>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>A view on the <code class="docutils literal notranslate"><span class="pre">tensor</span></code>. </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7IBuffer4viewE9SharedPtrNSt6size_tE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7IBuffer4viewE9SharedPtrNSt6size_tE"></span><span id="_CPPv2N12tensorrt_llm7runtime7IBuffer4viewE9SharedPtrNSt6size_tE"></span><span id="tensorrt_llm::runtime::IBuffer::view__SharedPtr.std::s"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1a58bc0fcaa0356cf9e20ce1d06b16c70f"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer9UniquePtrE" title="tensorrt_llm::runtime::IBuffer::UniquePtr"><span class="n"><span class="pre">UniquePtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">view</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer9SharedPtrE" title="tensorrt_llm::runtime::IBuffer::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tensor</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer4viewE9SharedPtrNSt6size_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Returns a view on the underlying <code class="docutils literal notranslate"><span class="pre">tensor</span></code> with a different size. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><ul class="simple">
|
||
<li><p><strong>tensor</strong> – The tensor to view. </p></li>
|
||
<li><p><strong>size</strong> – The size of the view. </p></li>
|
||
</ul>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>A view on the <code class="docutils literal notranslate"><span class="pre">tensor</span></code>. </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7IBuffer4viewE14UniqueConstPtrRR9TConstPtrNSt6size_tE">
|
||
<span id="_CPPv3I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7IBuffer4viewERR9TConstPtrNSt6size_tE"></span><span id="_CPPv2I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7IBuffer4viewERR9TConstPtrNSt6size_tE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TConstPtr</span></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">enable_if_t</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">is_const_v</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime18PointerElementTypeE" title="tensorrt_llm::runtime::PointerElementType"><span class="n"><span class="pre">PointerElementType</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7IBuffer4viewE14UniqueConstPtrRR9TConstPtrNSt6size_tE" title="tensorrt_llm::runtime::IBuffer::view::TConstPtr"><span class="n"><span class="pre">TConstPtr</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="kt"><span class="pre">int</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1af28ec3097ffc66614052cef9392265bb"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer14UniqueConstPtrE" title="tensorrt_llm::runtime::IBuffer::UniqueConstPtr"><span class="n"><span class="pre">UniqueConstPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">view</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7IBuffer4viewE14UniqueConstPtrRR9TConstPtrNSt6size_tE" title="tensorrt_llm::runtime::IBuffer::view::TConstPtr"><span class="n"><span class="pre">TConstPtr</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">tensor</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0_NSt11enable_if_tINSt10is_const_vI18PointerElementTypeI9TConstPtrEEEiEEEN12tensorrt_llm7runtime7IBuffer4viewE14UniqueConstPtrRR9TConstPtrNSt6size_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7IBuffer4wrapEPv8DataTypeNSt6size_tENSt6size_tE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7IBuffer4wrapEPv8DataTypeNSt6size_tENSt6size_tE"></span><span id="_CPPv2N12tensorrt_llm7runtime7IBuffer4wrapEPv8DataTypeNSt6size_tENSt6size_tE"></span><span id="tensorrt_llm::runtime::IBuffer::wrap__voidP.DataType.std::s.std::s"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1a6cf1157eb353c90dbe12711be9af63f2"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer9UniquePtrE" title="tensorrt_llm::runtime::IBuffer::UniquePtr"><span class="n"><span class="pre">UniquePtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">wrap</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="n sig-param"><span class="pre">data</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer8DataTypeE" title="tensorrt_llm::runtime::IBuffer::DataType"><span class="n"><span class="pre">DataType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">type</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">capacity</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer4wrapEPv8DataTypeNSt6size_tENSt6size_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Wraps the given <code class="docutils literal notranslate"><span class="pre">data</span></code> in an <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1IBuffer"><span class="std std-ref"><span class="pre">IBuffer</span></span></a></code>. The <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1IBuffer"><span class="std std-ref"><span class="pre">IBuffer</span></span></a></code> will not own the underlying <code class="docutils literal notranslate"><span class="pre">data</span></code> and cannot be resized beyond <code class="docutils literal notranslate"><span class="pre">capacity</span></code>. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><ul class="simple">
|
||
<li><p><strong>data</strong> – The data to wrap. </p></li>
|
||
<li><p><strong>type</strong> – The data type of the <code class="docutils literal notranslate"><span class="pre">data</span></code>. </p></li>
|
||
<li><p><strong>size</strong> – The size of the buffer. </p></li>
|
||
<li><p><strong>capacity</strong> – The capacity of the buffer. </p></li>
|
||
</ul>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>An <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1IBuffer"><span class="std std-ref"><span class="pre">IBuffer</span></span></a></code>. </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7IBuffer4wrapEPv8DataTypeNSt6size_tE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7IBuffer4wrapEPv8DataTypeNSt6size_tE"></span><span id="_CPPv2N12tensorrt_llm7runtime7IBuffer4wrapEPv8DataTypeNSt6size_tE"></span><span id="tensorrt_llm::runtime::IBuffer::wrap__voidP.DataType.std::s"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1af1fb9e5497bfc31b149ff9477161068d"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer9UniquePtrE" title="tensorrt_llm::runtime::IBuffer::UniquePtr"><span class="n"><span class="pre">UniquePtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">wrap</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="n sig-param"><span class="pre">data</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer8DataTypeE" title="tensorrt_llm::runtime::IBuffer::DataType"><span class="n"><span class="pre">DataType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">type</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer4wrapEPv8DataTypeNSt6size_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime7IBuffer4wrapE9UniquePtrP1TNSt6size_tENSt6size_tE">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime7IBuffer4wrapEP1TNSt6size_tENSt6size_tE"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime7IBuffer4wrapEP1TNSt6size_tENSt6size_tE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1a1f5dd20e613af0bc2a05f0b09343535d"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer9UniquePtrE" title="tensorrt_llm::runtime::IBuffer::UniquePtr"><span class="n"><span class="pre">UniquePtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">wrap</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime7IBuffer4wrapE9UniquePtrP1TNSt6size_tENSt6size_tE" title="tensorrt_llm::runtime::IBuffer::wrap::T"><span class="n"><span class="pre">T</span></span></a><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="n sig-param"><span class="pre">data</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">capacity</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime7IBuffer4wrapE9UniquePtrP1TNSt6size_tENSt6size_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime7IBuffer4wrapE9UniquePtrP1TNSt6size_tE">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime7IBuffer4wrapEP1TNSt6size_tE"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime7IBuffer4wrapEP1TNSt6size_tE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1a2e4240d0d7f78a278716c8faccf5e9f9"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer9UniquePtrE" title="tensorrt_llm::runtime::IBuffer::UniquePtr"><span class="n"><span class="pre">UniquePtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">wrap</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime7IBuffer4wrapE9UniquePtrP1TNSt6size_tE" title="tensorrt_llm::runtime::IBuffer::wrap::T"><span class="n"><span class="pre">T</span></span></a><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="n sig-param"><span class="pre">data</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime7IBuffer4wrapE9UniquePtrP1TNSt6size_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime7IBuffer4wrapE9UniquePtrRNSt6vectorI1TEE">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime7IBuffer4wrapERNSt6vectorI1TEE"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime7IBuffer4wrapERNSt6vectorI1TEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1a84fdf2b484eee3440646edfece5b85b0"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer9UniquePtrE" title="tensorrt_llm::runtime::IBuffer::UniquePtr"><span class="n"><span class="pre">UniquePtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">wrap</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime7IBuffer4wrapE9UniquePtrRNSt6vectorI1TEE" title="tensorrt_llm::runtime::IBuffer::wrap::T"><span class="n"><span class="pre">T</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">v</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime7IBuffer4wrapE9UniquePtrRNSt6vectorI1TEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7IBuffer10memoryTypeEPKv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7IBuffer10memoryTypeEPKv"></span><span id="_CPPv2N12tensorrt_llm7runtime7IBuffer10memoryTypeEPKv"></span><span id="tensorrt_llm::runtime::IBuffer::memoryType__voidCP"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1a2c714d66d0336cac0708008e59dd71cd"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="tensorrt_llm::runtime::MemoryType"><span class="n"><span class="pre">MemoryType</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">memoryType</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="n sig-param"><span class="pre">data</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer10memoryTypeEPKv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Determine the memory type of a pointer. </p>
|
||
</dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-protected-functions">Protected Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7IBuffer7IBufferEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7IBuffer7IBufferEv"></span><span id="_CPPv2N12tensorrt_llm7runtime7IBuffer7IBufferEv"></span><span id="tensorrt_llm::runtime::IBuffer::IBuffer"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1ad6ad0acb873c47b64fc8fec029696ac2"></span><span class="sig-name descname"><span class="n"><span class="pre">IBuffer</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">default</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7IBuffer7IBufferEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime7IBuffer7toBytesENSt6size_tE">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime7IBuffer7toBytesENSt6size_tE"></span><span id="_CPPv2NK12tensorrt_llm7runtime7IBuffer7toBytesENSt6size_tE"></span><span id="tensorrt_llm::runtime::IBuffer::toBytes__std::sC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IBuffer_1aade9d63190ac92fdb3561a61af5ab2ee"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">toBytes</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">size</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime7IBuffer7toBytesENSt6size_tE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Returns an array index or size in bytes. </p>
|
||
</dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I_10MemoryTypeEN12tensorrt_llm7runtime16MemoryTypeStringE">
|
||
<span id="_CPPv3I_10MemoryTypeEN12tensorrt_llm7runtime16MemoryTypeStringE"></span><span id="_CPPv2I_10MemoryTypeEN12tensorrt_llm7runtime16MemoryTypeStringE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10MemoryTypeE" title="tensorrt_llm::runtime::MemoryType"><span class="n"><span class="pre">MemoryType</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1MemoryTypeString"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">MemoryTypeString</span></span></span><a class="headerlink" href="#_CPPv4I_10MemoryTypeEN12tensorrt_llm7runtime16MemoryTypeStringE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1MemoryTypeString_3_01MemoryType_1_1kCPU_01_4"></span><span class="sig-name descname"><span class="pre">template<></span> <span class="pre">kCPU</span> <span class="pre">></span></span></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime19PhonyNameDueToError5valueE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime19PhonyNameDueToError5valueE"></span><span id="_CPPv2N12tensorrt_llm7runtime19PhonyNameDueToError5valueE"></span><span id="tensorrt_llm::runtime::PhonyNameDueToError::value__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1MemoryTypeString_3_01MemoryType_1_1kCPU_01_4_1af3b289e3cfd246c2b1ee25aec5eeb6e1"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"CPU"</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime19PhonyNameDueToError5valueE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1MemoryTypeString_3_01MemoryType_1_1kGPU_01_4"></span><span class="sig-name descname"><span class="pre">template<></span> <span class="pre">kGPU</span> <span class="pre">></span></span></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1MemoryTypeString_3_01MemoryType_1_1kGPU_01_4_1a0fc1c309e17dd52d719cafddbcc6cb12"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"GPU"</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1MemoryTypeString_3_01MemoryType_1_1kPINNED_01_4"></span><span class="sig-name descname"><span class="pre">template<></span> <span class="pre">kPINNED</span> <span class="pre">></span></span></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1MemoryTypeString_3_01MemoryType_1_1kPINNED_01_4_1a5e877eb90ca7cbd5b5269d4f092e69e4"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"PINNED"</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1MemoryTypeString_3_01MemoryType_1_1kPINNEDPOOL_01_4"></span><span class="sig-name descname"><span class="pre">template<></span> <span class="pre">kPINNEDPOOL</span> <span class="pre">></span></span></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1MemoryTypeString_3_01MemoryType_1_1kPINNEDPOOL_01_4_1aad3eee9ea444088033995a877a00e9c9"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"PINNEDPOOL"</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1MemoryTypeString_3_01MemoryType_1_1kUVM_01_4"></span><span class="sig-name descname"><span class="pre">template<></span> <span class="pre">kUVM</span> <span class="pre">></span></span></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1MemoryTypeString_3_01MemoryType_1_1kUVM_01_4_1a743212e7128317c3b6a0af4ad619afa9"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="s"><span class="pre">"UVM"</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0_bEN12tensorrt_llm7runtime11TRTDataTypeE">
|
||
<span id="_CPPv3I0_bEN12tensorrt_llm7runtime11TRTDataTypeE"></span><span id="_CPPv2I0_bEN12tensorrt_llm7runtime11TRTDataTypeE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">false</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TRTDataType</span></span></span><a class="headerlink" href="#_CPPv4I0_bEN12tensorrt_llm7runtime11TRTDataTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="docutils container">
|
||
<em>#include <iBuffer.h></em></div>
|
||
<p>For converting a C++ data type to a TensorRT data type. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeIbEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime11TRTDataTypeIbEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime11TRTDataTypeIbEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01bool_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TRTDataType</span></span></span><span class="p"><span class="pre"><</span></span><span class="kt"><span class="pre">bool</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeIbEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11TRTDataTypeIbE5valueE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11TRTDataTypeIbE5valueE"></span><span id="_CPPv2N12tensorrt_llm7runtime11TRTDataTypeIbE5valueE"></span><span id="tensorrt_llm::runtime::TRTDataType:b:::value__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01bool_01_4_1a590c40a56a5eabf6069340edc07d84e4"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">kBOOL</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeIbE5valueE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeIfEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime11TRTDataTypeIfEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime11TRTDataTypeIfEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01float_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TRTDataType</span></span></span><span class="p"><span class="pre"><</span></span><span class="kt"><span class="pre">float</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeIfEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11TRTDataTypeIfE5valueE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11TRTDataTypeIfE5valueE"></span><span id="_CPPv2N12tensorrt_llm7runtime11TRTDataTypeIfE5valueE"></span><span id="tensorrt_llm::runtime::TRTDataType:float:::value__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01float_01_4_1a0a72c60c5a9b946e64a960d1075a5af9"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">kFLOAT</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeIfE5valueE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeI4halfEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime11TRTDataTypeI4halfEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime11TRTDataTypeI4halfEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01half_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TRTDataType</span></span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">half</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeI4halfEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11TRTDataTypeI4halfE5valueE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11TRTDataTypeI4halfE5valueE"></span><span id="_CPPv2N12tensorrt_llm7runtime11TRTDataTypeI4halfE5valueE"></span><span id="tensorrt_llm::runtime::TRTDataType:half:::value__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01half_01_4_1a2febf53869fcfd71830c674652e7c7d4"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">kHALF</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeI4halfE5valueE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01kernels_1_1FinishedState_01_4"></span><span class="sig-name descname"><span class="pre">template<></span> <span class="pre">FinishedState</span> <span class="pre">></span></span></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01kernels_1_1FinishedState_01_4_1af0e2893266f9a1af9a97644162b45afa"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4I0_bEN12tensorrt_llm7runtime11TRTDataTypeE" title="tensorrt_llm::runtime::TRTDataType"><span class="n"><span class="pre">TRTDataType</span></span></a><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">kernels</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">FinishedState</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">UnderlyingType</span></span><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">value</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01kernels_1_1KVCacheIndex_01_4"></span><span class="sig-name descname"><span class="pre">template<></span> <span class="pre">KVCacheIndex</span> <span class="pre">></span></span></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01kernels_1_1KVCacheIndex_01_4_1a8a9684f587b9b6ac46ac769f6f205419"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4I0_bEN12tensorrt_llm7runtime11TRTDataTypeE" title="tensorrt_llm::runtime::TRTDataType"><span class="n"><span class="pre">TRTDataType</span></span></a><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">kernels</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">KVCacheIndex</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">UnderlyingType</span></span><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">value</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01runtime_1_1RequestType_01_4"></span><span class="sig-name descname"><span class="pre">template<></span> <span class="pre">RequestType</span> <span class="pre">></span></span></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01runtime_1_1RequestType_01_4_1a689d61d98e3959c3f520274718c23541"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4I0_bEN12tensorrt_llm7runtime11TRTDataTypeE" title="tensorrt_llm::runtime::TRTDataType"><span class="n"><span class="pre">TRTDataType</span></span></a><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">underlying_type_t</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11RequestTypeE" title="tensorrt_llm::runtime::RequestType"><span class="n"><span class="pre">RequestType</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">value</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01std_1_1int32__t_01_4"></span><span class="sig-name descname"><span class="pre">template<></span> <span class="pre">int32_t</span> <span class="pre">></span></span></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01std_1_1int32__t_01_4_1a10403a956a1d979e0cd10c8034e9f3e1"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">kINT32</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01std_1_1int64__t_01_4"></span><span class="sig-name descname"><span class="pre">template<></span> <span class="pre">int64_t</span> <span class="pre">></span></span></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01std_1_1int64__t_01_4_1af2839adf78453826b3a7f8422c9130fb"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">kINT64</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01std_1_1int8__t_01_4"></span><span class="sig-name descname"><span class="pre">template<></span> <span class="pre">int8_t</span> <span class="pre">></span></span></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01std_1_1int8__t_01_4_1a996ff7940394a814ae164728312bf5ff"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">kINT8</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01std_1_1uint32__t_01_4"></span><span class="sig-name descname"><span class="pre">template<></span> <span class="pre">uint32_t</span> <span class="pre">></span></span></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01std_1_1uint32__t_01_4_1a96b97cf0edb4ad984dd69278acfabd10"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14BufferDataTypeE" title="tensorrt_llm::runtime::BufferDataType"><span class="n"><span class="pre">BufferDataType</span></span></a><span class="p"><span class="pre">{</span></span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">kINT32</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="k"><span class="pre">true</span></span><span class="p"><span class="pre">}</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01std_1_1uint64__t_01_4"></span><span class="sig-name descname"><span class="pre">template<></span> <span class="pre">uint64_t</span> <span class="pre">></span></span></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01std_1_1uint64__t_01_4_1a6d279398fbfedbbfffd0f7fbd1eb80b8"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14BufferDataTypeE" title="tensorrt_llm::runtime::BufferDataType"><span class="n"><span class="pre">BufferDataType</span></span></a><span class="p"><span class="pre">{</span></span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">kINT64</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="k"><span class="pre">true</span></span><span class="p"><span class="pre">}</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01std_1_1uint8__t_01_4"></span><span class="sig-name descname"><span class="pre">template<></span> <span class="pre">uint8_t</span> <span class="pre">></span></span></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01std_1_1uint8__t_01_4_1a60d418bc6cea9f845ee353d78f0c3d66"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">kUINT8</span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime11TRTDataTypeIP1TEE">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime11TRTDataTypeIP1TEE"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime11TRTDataTypeIP1TEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01T_01_5_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TRTDataType</span></span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime11TRTDataTypeIP1TEE" title="tensorrt_llm::runtime::TRTDataType<T*>::T"><span class="n"><span class="pre">T</span></span></a><span class="p"><span class="pre">*</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime11TRTDataTypeIP1TEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11TRTDataTypeIP1TE5valueE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11TRTDataTypeIP1TE5valueE"></span><span id="_CPPv2N12tensorrt_llm7runtime11TRTDataTypeIP1TE5valueE"></span><span id="tensorrt_llm::runtime::TRTDataType:TP:::value__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01T_01_5_01_4_1a6e605b21926aa5997834a5ed9069610d"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14BufferDataTypeE" title="tensorrt_llm::runtime::BufferDataType"><span class="n"><span class="pre">BufferDataType</span></span></a><span class="p"><span class="pre">{</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeIP1TE15kUnderlyingTypeE" title="tensorrt_llm::runtime::TRTDataType<T*>::kUnderlyingType"><span class="n"><span class="pre">kUnderlyingType</span></span></a><span class="p"><span class="pre">.</span></span><span class="n"><span class="pre">getDataType</span></span><span class="p"><span class="pre">(</span></span><span class="p"><span class="pre">)</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeIP1TE15kUnderlyingTypeE" title="tensorrt_llm::runtime::TRTDataType<T*>::kUnderlyingType"><span class="n"><span class="pre">kUnderlyingType</span></span></a><span class="p"><span class="pre">.</span></span><span class="n"><span class="pre">isUnsigned</span></span><span class="p"><span class="pre">(</span></span><span class="p"><span class="pre">)</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="k"><span class="pre">true</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeIP1TE5valueE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-static-attributes">Private Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11TRTDataTypeIP1TE15kUnderlyingTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11TRTDataTypeIP1TE15kUnderlyingTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11TRTDataTypeIP1TE15kUnderlyingTypeE"></span><span id="tensorrt_llm::runtime::TRTDataType:TP:::kUnderlyingType__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01T_01_5_01_4_1aa44a2359868876e32c25fdab6f9faf2d"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kUnderlyingType</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14BufferDataTypeE" title="tensorrt_llm::runtime::BufferDataType"><span class="n"><span class="pre">BufferDataType</span></span></a><span class="p"><span class="pre">{</span></span><span class="n"><span class="pre">TRTDataType</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">remove_const_t</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime11TRTDataTypeIP1TEE" title="tensorrt_llm::runtime::TRTDataType<T*>::T"><span class="n"><span class="pre">T</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="k"><span class="pre">false</span></span><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">value</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeIP1TE15kUnderlyingTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeIPvEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime11TRTDataTypeIPvEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime11TRTDataTypeIPvEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01void_01_5_01_4"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TRTDataType</span></span></span><span class="p"><span class="pre"><</span></span><span class="kt"><span class="pre">void</span></span><span class="p"><span class="pre">*</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4IEN12tensorrt_llm7runtime11TRTDataTypeIPvEE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11TRTDataTypeIPvE5valueE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11TRTDataTypeIPvE5valueE"></span><span id="_CPPv2N12tensorrt_llm7runtime11TRTDataTypeIPvE5valueE"></span><span id="tensorrt_llm::runtime::TRTDataType:voidP:::value__auto"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1TRTDataType_3_01void_01_5_01_4_1a63d38011d7c02b6ce399d457537b1a93"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="kt"><span class="pre">auto</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">value</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14BufferDataTypeE" title="tensorrt_llm::runtime::BufferDataType"><span class="n"><span class="pre">BufferDataType</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14BufferDataType15kTrtPointerTypeE" title="tensorrt_llm::runtime::BufferDataType::kTrtPointerType"><span class="n"><span class="pre">kTrtPointerType</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11TRTDataTypeIPvE5valueE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="igptdecoderbatched-h">
|
||
<h2>iGptDecoderBatched.h<a class="headerlink" href="#igptdecoderbatched-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1batch__manager"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">batch_manager</span></span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatchedE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime18IGptDecoderBatchedE"></span><span id="_CPPv2N12tensorrt_llm7runtime18IGptDecoderBatchedE"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">IGptDecoderBatched</span></span></span><span class="w"> </span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="k"><span class="pre">public</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoderE" title="tensorrt_llm::runtime::IStatefulGptDecoder"><span class="n"><span class="pre">IStatefulGptDecoder</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatchedE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="docutils container">
|
||
<em>#include <iGptDecoderBatched.h></em></div>
|
||
<p>GPT decoder class with support for in-flight batching. </p>
|
||
<p>Subclassed by <a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1GptDecoderBatched"><span class="std std-ref">tensorrt_llm::runtime::GptDecoderBatched</span></a></p>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched13CudaStreamPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime18IGptDecoderBatched13CudaStreamPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime18IGptDecoderBatched13CudaStreamPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1a2a27a496ac11aeb918dede4d513568aa"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">CudaStreamPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10CudaStreamE" title="tensorrt_llm::runtime::CudaStream"><span class="n"><span class="pre">CudaStream</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched13CudaStreamPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched13LlmRequestPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime18IGptDecoderBatched13LlmRequestPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime18IGptDecoderBatched13LlmRequestPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1a41f59ec19ac27bdc5cd92778f3d8d2a9"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">LlmRequestPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm13batch_managerE" title="tensorrt_llm::batch_manager"><span class="n"><span class="pre">batch_manager</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">LlmRequest</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched13LlmRequestPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched13RequestVectorE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime18IGptDecoderBatched13RequestVectorE"></span><span id="_CPPv2N12tensorrt_llm7runtime18IGptDecoderBatched13RequestVectorE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1ad580b2d9549986c709a235dc161f21c0"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">RequestVector</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched13LlmRequestPtrE" title="tensorrt_llm::runtime::IGptDecoderBatched::LlmRequestPtr"><span class="n"><span class="pre">LlmRequestPtr</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched13RequestVectorE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime18IGptDecoderBatched9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime18IGptDecoderBatched9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1a99e95a44eda53ca55f2e7efeba372229"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched9TensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched23DecoderFinishedEventPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime18IGptDecoderBatched23DecoderFinishedEventPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime18IGptDecoderBatched23DecoderFinishedEventPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1a01e9c005bc10e8caa062909d9e11c5c5"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">DecoderFinishedEventPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">unique_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch20DecoderFinishedEventE" title="tensorrt_llm::runtime::decoder_batch::DecoderFinishedEvent"><span class="n"><span class="pre">DecoderFinishedEvent</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched23DecoderFinishedEventPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched24setupExplicitDraftTokensEN26ExplicitDraftTokensBuffers6InputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime18IGptDecoderBatched24setupExplicitDraftTokensEN26ExplicitDraftTokensBuffers6InputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime18IGptDecoderBatched24setupExplicitDraftTokensEN26ExplicitDraftTokensBuffers6InputsE"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched::setupExplicitDraftTokens__ExplicitDraftTokensBuffers::Inputs"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1a1893ec85918a1d3710d0ae82d7c9c1c5"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setupExplicitDraftTokens</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffersE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers"><span class="n"><span class="pre">ExplicitDraftTokensBuffers</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers6InputsE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::Inputs"><span class="n"><span class="pre">Inputs</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">explicitDraftTokensBuffers</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched24setupExplicitDraftTokensEN26ExplicitDraftTokensBuffers6InputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Setup buffers for ExplicitDraftTokens decoding. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched10setupEagleEN12EagleBuffers6InputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime18IGptDecoderBatched10setupEagleEN12EagleBuffers6InputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime18IGptDecoderBatched10setupEagleEN12EagleBuffers6InputsE"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched::setupEagle__EagleBuffers::Inputs"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1a470d089a9032b6d1d942bb9fb108e086"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setupEagle</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffersE" title="tensorrt_llm::runtime::EagleBuffers"><span class="n"><span class="pre">EagleBuffers</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6InputsE" title="tensorrt_llm::runtime::EagleBuffers::Inputs"><span class="n"><span class="pre">Inputs</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">eagleBuffers</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched10setupEagleEN12EagleBuffers6InputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Setup buffers for Eagle decoding. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched14setupLookaheadE24LookaheadDecodingBuffers">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime18IGptDecoderBatched14setupLookaheadE24LookaheadDecodingBuffers"></span><span id="_CPPv2N12tensorrt_llm7runtime18IGptDecoderBatched14setupLookaheadE24LookaheadDecodingBuffers"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched::setupLookahead__LookaheadDecodingBuffers"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1aa42a0827f5ac6cda9197d945cc47d1a9"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setupLookahead</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffersE" title="tensorrt_llm::runtime::LookaheadDecodingBuffers"><span class="n"><span class="pre">LookaheadDecodingBuffers</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">lookaheadDecodingBuffers</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched14setupLookaheadE24LookaheadDecodingBuffers" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Setup buffers for Lookahead decoding. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched16disableLookaheadE10SizeType32RK13RequestVector">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime18IGptDecoderBatched16disableLookaheadE10SizeType32RK13RequestVector"></span><span id="_CPPv2N12tensorrt_llm7runtime18IGptDecoderBatched16disableLookaheadE10SizeType32RK13RequestVector"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched::disableLookahead__SizeType32.RequestVectorCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1ad7da8b95cd1e071e9d2cdf34b14031b4"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">disableLookahead</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxBatchSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched13RequestVectorE" title="tensorrt_llm::runtime::IGptDecoderBatched::RequestVector"><span class="n"><span class="pre">RequestVector</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">genRequests</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched16disableLookaheadE10SizeType32RK13RequestVector" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Disable Lookahead decoding. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched12forwardAsyncERN13decoder_batch6OutputERKN13decoder_batch5InputE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime18IGptDecoderBatched12forwardAsyncERN13decoder_batch6OutputERKN13decoder_batch5InputE"></span><span id="_CPPv2N12tensorrt_llm7runtime18IGptDecoderBatched12forwardAsyncERN13decoder_batch6OutputERKN13decoder_batch5InputE"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched::forwardAsync__decoder_batch::OutputR.decoder_batch::InputCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1ab528a9f6098123232f8b0393754f5c35"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched23DecoderFinishedEventPtrE" title="tensorrt_llm::runtime::IGptDecoderBatched::DecoderFinishedEventPtr"><span class="n"><span class="pre">DecoderFinishedEventPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">forwardAsync</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch6OutputE" title="tensorrt_llm::runtime::decoder_batch::Output"><span class="n"><span class="pre">Output</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">output</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5InputE" title="tensorrt_llm::runtime::decoder_batch::Input"><span class="n"><span class="pre">Input</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">input</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched12forwardAsyncERN13decoder_batch6OutputERKN13decoder_batch5InputE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Run one step for all requests without blocking the host process and return the token for synchronization. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched11forwardSyncERKN13decoder_batch20DecoderFinishedEventERN13decoder_batch6OutputERKN13decoder_batch5InputE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime18IGptDecoderBatched11forwardSyncERKN13decoder_batch20DecoderFinishedEventERN13decoder_batch6OutputERKN13decoder_batch5InputE"></span><span id="_CPPv2N12tensorrt_llm7runtime18IGptDecoderBatched11forwardSyncERKN13decoder_batch20DecoderFinishedEventERN13decoder_batch6OutputERKN13decoder_batch5InputE"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched::forwardSync__decoder_batch::DecoderFinishedEventCR.decoder_batch::OutputR.decoder_batch::InputCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1affc4dd41569723270cfe56f77af362cb"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">forwardSync</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch20DecoderFinishedEventE" title="tensorrt_llm::runtime::decoder_batch::DecoderFinishedEvent"><span class="n"><span class="pre">DecoderFinishedEvent</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">token</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch6OutputE" title="tensorrt_llm::runtime::decoder_batch::Output"><span class="n"><span class="pre">Output</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">output</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5InputE" title="tensorrt_llm::runtime::decoder_batch::Input"><span class="n"><span class="pre">Input</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">input</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched11forwardSyncERKN13decoder_batch20DecoderFinishedEventERN13decoder_batch6OutputERKN13decoder_batch5InputE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Call decoder forwardSync and wait for the call to <code class="docutils literal notranslate"><span class="pre">forwardAsync</span></code> associated with a token to complete. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched11forwardSyncERKN13decoder_batch20DecoderFinishedEventE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime18IGptDecoderBatched11forwardSyncERKN13decoder_batch20DecoderFinishedEventE"></span><span id="_CPPv2N12tensorrt_llm7runtime18IGptDecoderBatched11forwardSyncERKN13decoder_batch20DecoderFinishedEventE"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched::forwardSync__decoder_batch::DecoderFinishedEventCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1a67d3692b8f58468dfe9abc7ec3bc267c"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">forwardSync</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch20DecoderFinishedEventE" title="tensorrt_llm::runtime::decoder_batch::DecoderFinishedEvent"><span class="n"><span class="pre">DecoderFinishedEvent</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">token</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched11forwardSyncERKN13decoder_batch20DecoderFinishedEventE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Wait for the call to <code class="docutils literal notranslate"><span class="pre">forwardAsync</span></code> associated with a token to complete. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched7forwardERN13decoder_batch6OutputERKN13decoder_batch5InputE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime18IGptDecoderBatched7forwardERN13decoder_batch6OutputERKN13decoder_batch5InputE"></span><span id="_CPPv2N12tensorrt_llm7runtime18IGptDecoderBatched7forwardERN13decoder_batch6OutputERKN13decoder_batch5InputE"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched::forward__decoder_batch::OutputR.decoder_batch::InputCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1a75a4af93b82153c19363ff55ceba5c74"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">forward</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch6OutputE" title="tensorrt_llm::runtime::decoder_batch::Output"><span class="n"><span class="pre">Output</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">output</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5InputE" title="tensorrt_llm::runtime::decoder_batch::Input"><span class="n"><span class="pre">Input</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">input</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched7forwardERN13decoder_batch6OutputERKN13decoder_batch5InputE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Run one step for all requests and wait for completion on the host. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched6getIdsE10SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime18IGptDecoderBatched6getIdsE10SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime18IGptDecoderBatched6getIdsE10SizeType32"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched::getIds__SizeType32C"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1a33761c3b92894dcf0c90c99eec594f48"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::IGptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getIds</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchIdx</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched6getIdsE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>batchIdx</strong> – index of the batch </p>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>[maxBeamWidth, maxInputLength + maxNewTokens], contains input token ids and generated token ids without padding for request <code class="docutils literal notranslate"><span class="pre">batchIdx</span></code>, on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched14getGatheredIdsE10SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime18IGptDecoderBatched14getGatheredIdsE10SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime18IGptDecoderBatched14getGatheredIdsE10SizeType32"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched::getGatheredIds__SizeType32C"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1a87aea58eba8d5daf55d05f74b4272c9f"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::IGptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getGatheredIds</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchIdx</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched14getGatheredIdsE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[batchSize, maxBeamWidth, maxInputLength + maxNewTokens], only used for beam search in <a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1GptDecoderBatched"><span class="std std-ref">GptDecoderBatched</span></a> It contains gathered token ids without padding, on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched8finalizeE10SizeType32RK14SamplingConfigb">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime18IGptDecoderBatched8finalizeE10SizeType32RK14SamplingConfigb"></span><span id="_CPPv2NK12tensorrt_llm7runtime18IGptDecoderBatched8finalizeE10SizeType32RK14SamplingConfigb"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched::finalize__SizeType32.SamplingConfigCR.bC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1a84922a0c24b92f88fc8daa135f0b0350"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEventE" title="tensorrt_llm::runtime::CudaEvent"><span class="n"><span class="pre">CudaEvent</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">finalize</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchIdx</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfigE" title="tensorrt_llm::runtime::SamplingConfig"><span class="n"><span class="pre">SamplingConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">samplingConfig</span></span>, <span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">streaming</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched8finalizeE10SizeType32RK14SamplingConfigb" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Gather final beam search results for request <code class="docutils literal notranslate"><span class="pre">batchIdx</span></code>. Result will only be available after event returned. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched11getFinishedEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime18IGptDecoderBatched11getFinishedEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime18IGptDecoderBatched11getFinishedEv"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched::getFinishedC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1a52b390b89507ba8ff7522e4842734a25"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="kt"><span class="pre">bool</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getFinished</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched11getFinishedEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[batchSize (actual)], marks finished requests (per batch) </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched16getFinishReasonsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime18IGptDecoderBatched16getFinishReasonsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime18IGptDecoderBatched16getFinishReasonsEv"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched::getFinishReasonsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1af36f79df9964e2a47ec8ce772ff3ce87"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::IGptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getFinishReasons</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched16getFinishReasonsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[batchSize, beamWidth], FinishedState value, on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched14getCumLogProbsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime18IGptDecoderBatched14getCumLogProbsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime18IGptDecoderBatched14getCumLogProbsEv"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched::getCumLogProbsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1abd25675ddaad98f83de1dd7583967003"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::IGptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getCumLogProbs</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched14getCumLogProbsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[batchSize, beamWidth], cumulative log probabilities (per beam), on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched14getCumLogProbsE10SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime18IGptDecoderBatched14getCumLogProbsE10SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime18IGptDecoderBatched14getCumLogProbsE10SizeType32"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched::getCumLogProbs__SizeType32C"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1aa439edf7974ba6ee97da980284b274cb"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::IGptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getCumLogProbs</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchIdx</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched14getCumLogProbsE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[beamWidth], cumulative log probabilities (per beam) for request batchIdx, on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched11getLogProbsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime18IGptDecoderBatched11getLogProbsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime18IGptDecoderBatched11getLogProbsEv"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched::getLogProbsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1a078bfb1bea3ab9ce5de3ec591831fd75"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::IGptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getLogProbs</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched11getLogProbsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[batchSize, beamWidth, maxSeqLen], log probabilities (per beam), on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched11getLogProbsE10SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime18IGptDecoderBatched11getLogProbsE10SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime18IGptDecoderBatched11getLogProbsE10SizeType32"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched::getLogProbs__SizeType32C"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1a6605a635a464f2c10b70a228a6b8c439"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::IGptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getLogProbs</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchIdx</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched11getLogProbsE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[beamWidth, maxSeqLen], cumulative log probabilities (per beam) for request batchIdx, on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched12getParentIdsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime18IGptDecoderBatched12getParentIdsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime18IGptDecoderBatched12getParentIdsEv"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched::getParentIdsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1ae3363796203488278b422b1622d3e4ad"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::IGptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getParentIds</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched12getParentIdsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched10getNbStepsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime18IGptDecoderBatched10getNbStepsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime18IGptDecoderBatched10getNbStepsEv"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched::getNbStepsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1a5976f401aa0265078c0fce4022343064"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getNbSteps</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched10getNbStepsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched15getDecodingModeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime18IGptDecoderBatched15getDecodingModeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime18IGptDecoderBatched15getDecodingModeEv"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched::getDecodingModeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1a33b8c888dcc1ace79bda8d2008e73818"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executorE" title="tensorrt_llm::executor"><span class="n"><span class="pre">executor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executor12DecodingModeE" title="tensorrt_llm::executor::DecodingMode"><span class="n"><span class="pre">DecodingMode</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getDecodingMode</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched15getDecodingModeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched11newRequestsERKNSt6vectorI10SizeType32EERKNSt6vectorIN13decoder_batch7RequestEEERKNSt6vectorI14SamplingConfigEERK11ModelConfig">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime18IGptDecoderBatched11newRequestsERKNSt6vectorI10SizeType32EERKNSt6vectorIN13decoder_batch7RequestEEERKNSt6vectorI14SamplingConfigEERK11ModelConfig"></span><span id="_CPPv2N12tensorrt_llm7runtime18IGptDecoderBatched11newRequestsERKNSt6vectorI10SizeType32EERKNSt6vectorIN13decoder_batch7RequestEEERKNSt6vectorI14SamplingConfigEERK11ModelConfig"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched::newRequests__std::vector:SizeType32:CR.std::vector:decoder_batch::Request:CR.std::vector:SamplingConfig:CR.ModelConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1a3a366eae2f23641969949be7da0b18be"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">newRequests</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">seqSlots</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batchE" title="tensorrt_llm::runtime::decoder_batch"><span class="n"><span class="pre">decoder_batch</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7RequestE" title="tensorrt_llm::runtime::decoder_batch::Request"><span class="n"><span class="pre">Request</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">requests</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfigE" title="tensorrt_llm::runtime::SamplingConfig"><span class="n"><span class="pre">SamplingConfig</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">samplingConfigs</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE" title="tensorrt_llm::runtime::ModelConfig"><span class="n"><span class="pre">ModelConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelConfig</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched11newRequestsERKNSt6vectorI10SizeType32EERKNSt6vectorIN13decoder_batch7RequestEEERKNSt6vectorI14SamplingConfigEERK11ModelConfig" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Initialize batched decoder at seqSlots with a new <code class="docutils literal notranslate"><span class="pre">requests</span></code>. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched18getNextDraftTokensEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime18IGptDecoderBatched18getNextDraftTokensEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime18IGptDecoderBatched18getNextDraftTokensEv"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched::getNextDraftTokensC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1a0b71e5d13e1e38331ffecd1dda3bfffc"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::IGptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getNextDraftTokens</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched18getNextDraftTokensEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[batchSize, maxTokensPerStep-1], predicted draft tokens for next step, on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched25getPrevDraftTokensLengthsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime18IGptDecoderBatched25getPrevDraftTokensLengthsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime18IGptDecoderBatched25getPrevDraftTokensLengthsEv"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched::getPrevDraftTokensLengthsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1a4dd436f1587e35f2c25c2d23fd0bd038"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::IGptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getPrevDraftTokensLengths</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched25getPrevDraftTokensLengthsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[batchSize], predicted draft tokens lengths for previous step, on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched25getNextDraftTokensLengthsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime18IGptDecoderBatched25getNextDraftTokensLengthsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime18IGptDecoderBatched25getNextDraftTokensLengthsEv"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched::getNextDraftTokensLengthsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1a9deab2068ba63616ed22f3e4b6a6cb68"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::IGptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getNextDraftTokensLengths</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched25getNextDraftTokensLengthsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[batchSize], predicted draft tokens lengths for next step, on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched24getAcceptedLengthsCumSumEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime18IGptDecoderBatched24getAcceptedLengthsCumSumEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime18IGptDecoderBatched24getAcceptedLengthsCumSumEv"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched::getAcceptedLengthsCumSumC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1af0119119328073e546adfa6200fbf3b0"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::IGptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getAcceptedLengthsCumSum</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched24getAcceptedLengthsCumSumEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[batchSize + 1], exclusive sum of accepted draft token lengths, on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched22getAcceptedPackedPathsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime18IGptDecoderBatched22getAcceptedPackedPathsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime18IGptDecoderBatched22getAcceptedPackedPathsEv"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched::getAcceptedPackedPathsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1adcfdee7f43d6215714554cb48db0e0d0"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched9TensorPtrE" title="tensorrt_llm::runtime::IGptDecoderBatched::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getAcceptedPackedPaths</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime18IGptDecoderBatched22getAcceptedPackedPathsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[batchSize, maxAcceptedDraftTokensPerStep], accepted paths packed into continuous tensor, on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-protected-functions">Protected Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched18IGptDecoderBatchedEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime18IGptDecoderBatched18IGptDecoderBatchedEv"></span><span id="_CPPv2N12tensorrt_llm7runtime18IGptDecoderBatched18IGptDecoderBatchedEv"></span><span id="tensorrt_llm::runtime::IGptDecoderBatched::IGptDecoderBatched"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched_1ab7ee2bc18a2287c62d86ebf02f2c6f68"></span><span class="sig-name descname"><span class="n"><span class="pre">IGptDecoderBatched</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">default</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime18IGptDecoderBatched18IGptDecoderBatchedEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime_1_1decoder__batch"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">decoder_batch</span></span></span><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-typedefs">Typedefs</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch6OutputE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch6OutputE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch6OutputE"></span><span class="target" id="iGptDecoderBatched_8h_1a575658a5bf5a41310da470e54bf965d9"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Output</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7decoderE" title="tensorrt_llm::runtime::decoder"><span class="n"><span class="pre">decoder</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7decoder6OutputE" title="tensorrt_llm::runtime::decoder::Output"><span class="n"><span class="pre">Output</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch6OutputE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch20DecoderFinishedEventE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch20DecoderFinishedEventE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch20DecoderFinishedEventE"></span><span id="tensorrt_llm::runtime::decoder_batch::DecoderFinishedEvent"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1DecoderFinishedEvent"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">DecoderFinishedEvent</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch20DecoderFinishedEventE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch20DecoderFinishedEvent20DecoderFinishedEventERR9CudaEventRKNSt6vectorIbEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch20DecoderFinishedEvent20DecoderFinishedEventERR9CudaEventRKNSt6vectorIbEE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch20DecoderFinishedEvent20DecoderFinishedEventERR9CudaEventRKNSt6vectorIbEE"></span><span id="tensorrt_llm::runtime::decoder_batch::DecoderFinishedEvent::DecoderFinishedEvent__CudaEventRR.std::vector:b:CR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1DecoderFinishedEvent_1a7107c8610559c2b48aca787f98f12468"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">DecoderFinishedEvent</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEventE" title="tensorrt_llm::runtime::CudaEvent"><span class="n"><span class="pre">CudaEvent</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">event</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="kt"><span class="pre">bool</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">active</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch20DecoderFinishedEvent20DecoderFinishedEventERR9CudaEventRKNSt6vectorIbEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch20DecoderFinishedEvent5eventE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch20DecoderFinishedEvent5eventE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch20DecoderFinishedEvent5eventE"></span><span id="tensorrt_llm::runtime::decoder_batch::DecoderFinishedEvent::event__CudaEvent"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1DecoderFinishedEvent_1a6be64a5eb91d76858c970de62c14d3dc"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime9CudaEventE" title="tensorrt_llm::runtime::CudaEvent"><span class="n"><span class="pre">CudaEvent</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">event</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch20DecoderFinishedEvent5eventE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch20DecoderFinishedEvent6activeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch20DecoderFinishedEvent6activeE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch20DecoderFinishedEvent6activeE"></span><span id="tensorrt_llm::runtime::decoder_batch::DecoderFinishedEvent::active__std::vector:b:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1DecoderFinishedEvent_1a65ca7d9e351e52591b974927642997ca"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="kt"><span class="pre">bool</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">active</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch20DecoderFinishedEvent6activeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch5InputE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch5InputE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch5InputE"></span><span id="tensorrt_llm::runtime::decoder_batch::Input"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Input"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Input</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5InputE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input14TensorConstPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch5Input14TensorConstPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch5Input14TensorConstPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Input_1aa8152e055448762bd78ad70f53eda8ba"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorConstPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor14SharedConstPtrE" title="tensorrt_llm::runtime::ITensor::SharedConstPtr"><span class="n"><span class="pre">SharedConstPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input14TensorConstPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch5Input9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch5Input9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Input_1ac182ee5aa51be63dfe20586ecaf40043"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input9TensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input5InputERKNSt6vectorI9TensorPtrEERKNSt6vectorIbEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch5Input5InputERKNSt6vectorI9TensorPtrEERKNSt6vectorIbEE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch5Input5InputERKNSt6vectorI9TensorPtrEERKNSt6vectorIbEE"></span><span id="tensorrt_llm::runtime::decoder_batch::Input::Input__std::vector:TensorPtr:CR.std::vector:b:CR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Input_1ad2b3fc875250c83be03032f984154830"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Input</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input9TensorPtrE" title="tensorrt_llm::runtime::decoder_batch::Input::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">logits</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="kt"><span class="pre">bool</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">active</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input5InputERKNSt6vectorI9TensorPtrEERKNSt6vectorIbEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input5InputERKNSt6vectorI9TensorPtrEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch5Input5InputERKNSt6vectorI9TensorPtrEE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch5Input5InputERKNSt6vectorI9TensorPtrEE"></span><span id="tensorrt_llm::runtime::decoder_batch::Input::Input__std::vector:TensorPtr:CR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Input_1ad5c9b6c5a35df62813d6bedaae41660f"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Input</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input9TensorPtrE" title="tensorrt_llm::runtime::decoder_batch::Input::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">logits</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input5InputERKNSt6vectorI9TensorPtrEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input6logitsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch5Input6logitsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch5Input6logitsE"></span><span id="tensorrt_llm::runtime::decoder_batch::Input::logits__std::vector:TensorPtr:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Input_1a8ba3d8d2f0b70d2edc27ce997a4af07a"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input9TensorPtrE" title="tensorrt_llm::runtime::decoder_batch::Input::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">logits</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input6logitsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input6activeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch5Input6activeE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch5Input6activeE"></span><span id="tensorrt_llm::runtime::decoder_batch::Input::active__std::vector:b:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Input_1a85bb8bda589fcebf0c19994d76a26968"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="kt"><span class="pre">bool</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">active</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input6activeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input16cacheIndirectionE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch5Input16cacheIndirectionE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch5Input16cacheIndirectionE"></span><span id="tensorrt_llm::runtime::decoder_batch::Input::cacheIndirection__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Input_1ad1b93f5e2145ae272a72811679f7673a"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input9TensorPtrE" title="tensorrt_llm::runtime::decoder_batch::Input::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">cacheIndirection</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input16cacheIndirectionE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input20predictedDraftLogitsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch5Input20predictedDraftLogitsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch5Input20predictedDraftLogitsE"></span><span id="tensorrt_llm::runtime::decoder_batch::Input::predictedDraftLogits__std::vector:std::vector:TensorPtr::"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Input_1a6692df8763ceda9c12f94f4553d32ed1"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input9TensorPtrE" title="tensorrt_llm::runtime::decoder_batch::Input::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">predictedDraftLogits</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input20predictedDraftLogitsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input8seqSlotsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch5Input8seqSlotsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch5Input8seqSlotsE"></span><span id="tensorrt_llm::runtime::decoder_batch::Input::seqSlots__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Input_1aaead2419bc0e8f3acc607c5936a13c6a"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input9TensorPtrE" title="tensorrt_llm::runtime::decoder_batch::Input::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">seqSlots</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input8seqSlotsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input25explicitDraftTokensInputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch5Input25explicitDraftTokensInputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch5Input25explicitDraftTokensInputsE"></span><span id="tensorrt_llm::runtime::decoder_batch::Input::explicitDraftTokensInputs__std::optional:ExplicitDraftTokensBuffers::EngineOutputs:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Input_1ae58255408d173c03cb6efb2a988b8479"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffersE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers"><span class="n"><span class="pre">ExplicitDraftTokensBuffers</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers13EngineOutputsE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::EngineOutputs"><span class="n"><span class="pre">EngineOutputs</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">explicitDraftTokensInputs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input25explicitDraftTokensInputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input29explicitDraftTokensLastInputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch5Input29explicitDraftTokensLastInputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch5Input29explicitDraftTokensLastInputsE"></span><span id="tensorrt_llm::runtime::decoder_batch::Input::explicitDraftTokensLastInputs__std::optional:ExplicitDraftTokensBuffers::EngineInputs:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Input_1a41a98f4e5c7f05757a82300ede64ac4d"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffersE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers"><span class="n"><span class="pre">ExplicitDraftTokensBuffers</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime26ExplicitDraftTokensBuffers12EngineInputsE" title="tensorrt_llm::runtime::ExplicitDraftTokensBuffers::EngineInputs"><span class="n"><span class="pre">EngineInputs</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">explicitDraftTokensLastInputs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input29explicitDraftTokensLastInputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input11eagleInputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch5Input11eagleInputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch5Input11eagleInputsE"></span><span id="tensorrt_llm::runtime::decoder_batch::Input::eagleInputs__std::optional:EagleBuffers::EngineOutputs:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Input_1ab53513121aa63b63d84569172619b207"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffersE" title="tensorrt_llm::runtime::EagleBuffers"><span class="n"><span class="pre">EagleBuffers</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers13EngineOutputsE" title="tensorrt_llm::runtime::EagleBuffers::EngineOutputs"><span class="n"><span class="pre">EngineOutputs</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">eagleInputs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input11eagleInputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input15eagleLastInputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch5Input15eagleLastInputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch5Input15eagleLastInputsE"></span><span id="tensorrt_llm::runtime::decoder_batch::Input::eagleLastInputs__std::optional:EagleBuffers::Inputs:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Input_1a68e11bed66f250eedc5e74d45973fcaf"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffersE" title="tensorrt_llm::runtime::EagleBuffers"><span class="n"><span class="pre">EagleBuffers</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime12EagleBuffers6InputsE" title="tensorrt_llm::runtime::EagleBuffers::Inputs"><span class="n"><span class="pre">Inputs</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">eagleLastInputs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input15eagleLastInputsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="istatefulgptdecoder-h">
|
||
<h2>iStatefulGptDecoder.h<a class="headerlink" href="#istatefulgptdecoder-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1batch__manager"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">batch_manager</span></span></span><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoderE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime19IStatefulGptDecoderE"></span><span id="_CPPv2N12tensorrt_llm7runtime19IStatefulGptDecoderE"></span><span id="tensorrt_llm::runtime::IStatefulGptDecoder"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IStatefulGptDecoder"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">IStatefulGptDecoder</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoderE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="docutils container">
|
||
<em>#include <iStatefulGptDecoder.h></em></div>
|
||
<p>GPT decoder class with support for in-flight batching. </p>
|
||
<p>Subclassed by <a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1IGptDecoderBatched"><span class="std std-ref">tensorrt_llm::runtime::IGptDecoderBatched</span></a></p>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder13CudaStreamPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime19IStatefulGptDecoder13CudaStreamPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime19IStatefulGptDecoder13CudaStreamPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IStatefulGptDecoder_1a310eb1dafd035c945176ea533ec74588"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">CudaStreamPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10CudaStreamE" title="tensorrt_llm::runtime::CudaStream"><span class="n"><span class="pre">CudaStream</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder13CudaStreamPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime19IStatefulGptDecoder9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime19IStatefulGptDecoder9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IStatefulGptDecoder_1a3987d8ce3a11babf865981af9d4ea18e"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder9TensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder5setupERKN8executor12DecodingModeE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType3210SizeType32N8nvinfer18DataTypeERK11ModelConfig">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime19IStatefulGptDecoder5setupERKN8executor12DecodingModeE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType3210SizeType32N8nvinfer18DataTypeERK11ModelConfig"></span><span id="_CPPv2N12tensorrt_llm7runtime19IStatefulGptDecoder5setupERKN8executor12DecodingModeE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType3210SizeType32N8nvinfer18DataTypeERK11ModelConfig"></span><span id="tensorrt_llm::runtime::IStatefulGptDecoder::setup__executor::DecodingModeCR.SizeType32.SizeType32.SizeType32.SizeType32.SizeType32.SizeType32.nvinfer1::DataType.ModelConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IStatefulGptDecoder_1a841d2576e3abba56e8880ad82ac1e763"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setup</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executorE" title="tensorrt_llm::executor"><span class="n"><span class="pre">executor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executor12DecodingModeE" title="tensorrt_llm::executor::DecodingMode"><span class="n"><span class="pre">DecodingMode</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">mode</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxBatchSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxBeamWidth</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxAttentionWindow</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">sinkTokenLength</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxSequenceLength</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxTokensPerStep</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">dtype</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE" title="tensorrt_llm::runtime::ModelConfig"><span class="n"><span class="pre">ModelConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelConfig</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder5setupERKN8executor12DecodingModeE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType3210SizeType32N8nvinfer18DataTypeERK11ModelConfig" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Setup the decoder before calling <code class="docutils literal notranslate"><a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1IStatefulGptDecoder_1a152618c3967cc29d314bd77d17c51e27"><span class="std std-ref"><span class="pre">forward()</span></span></a></code>, also calls reshapeBuffers. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder8newBatchERK15GenerationInputRK16GenerationOutputRK14SamplingConfigRK11ModelConfig">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime19IStatefulGptDecoder8newBatchERK15GenerationInputRK16GenerationOutputRK14SamplingConfigRK11ModelConfig"></span><span id="_CPPv2N12tensorrt_llm7runtime19IStatefulGptDecoder8newBatchERK15GenerationInputRK16GenerationOutputRK14SamplingConfigRK11ModelConfig"></span><span id="tensorrt_llm::runtime::IStatefulGptDecoder::newBatch__GenerationInputCR.GenerationOutputCR.SamplingConfigCR.ModelConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IStatefulGptDecoder_1a013c05675816f9abaf383398cf0fed77"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">newBatch</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15GenerationInputE" title="tensorrt_llm::runtime::GenerationInput"><span class="n"><span class="pre">GenerationInput</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">inputs</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16GenerationOutputE" title="tensorrt_llm::runtime::GenerationOutput"><span class="n"><span class="pre">GenerationOutput</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">outputs</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfigE" title="tensorrt_llm::runtime::SamplingConfig"><span class="n"><span class="pre">SamplingConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">samplingConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE" title="tensorrt_llm::runtime::ModelConfig"><span class="n"><span class="pre">ModelConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelConfig</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder8newBatchERK15GenerationInputRK16GenerationOutputRK14SamplingConfigRK11ModelConfig" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Initialize the decoder with new batch of inputs. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder12forwardAsyncERN7decoder6OutputERKN7decoder5InputE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime19IStatefulGptDecoder12forwardAsyncERN7decoder6OutputERKN7decoder5InputE"></span><span id="_CPPv2N12tensorrt_llm7runtime19IStatefulGptDecoder12forwardAsyncERN7decoder6OutputERKN7decoder5InputE"></span><span id="tensorrt_llm::runtime::IStatefulGptDecoder::forwardAsync__decoder::OutputR.decoder::InputCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IStatefulGptDecoder_1a7bb51780e6e9a6c41de214ffac8b07be"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">forwardAsync</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7decoderE" title="tensorrt_llm::runtime::decoder"><span class="n"><span class="pre">decoder</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7decoder6OutputE" title="tensorrt_llm::runtime::decoder::Output"><span class="n"><span class="pre">Output</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">output</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7decoderE" title="tensorrt_llm::runtime::decoder"><span class="n"><span class="pre">decoder</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7decoder5InputE" title="tensorrt_llm::runtime::decoder::Input"><span class="n"><span class="pre">Input</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">input</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder12forwardAsyncERN7decoder6OutputERKN7decoder5InputE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Run one step for all requests without blocking the host thread. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder11forwardSyncEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime19IStatefulGptDecoder11forwardSyncEv"></span><span id="_CPPv2N12tensorrt_llm7runtime19IStatefulGptDecoder11forwardSyncEv"></span><span id="tensorrt_llm::runtime::IStatefulGptDecoder::forwardSync"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IStatefulGptDecoder_1a2a2a7296d9137e434da174a04626722a"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">forwardSync</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder11forwardSyncEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Wait for the last call to <code class="docutils literal notranslate"><span class="pre">forwardAsync</span></code> to complete. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder7forwardERN7decoder6OutputERKN7decoder5InputE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime19IStatefulGptDecoder7forwardERN7decoder6OutputERKN7decoder5InputE"></span><span id="_CPPv2N12tensorrt_llm7runtime19IStatefulGptDecoder7forwardERN7decoder6OutputERKN7decoder5InputE"></span><span id="tensorrt_llm::runtime::IStatefulGptDecoder::forward__decoder::OutputR.decoder::InputCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IStatefulGptDecoder_1a152618c3967cc29d314bd77d17c51e27"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">forward</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7decoderE" title="tensorrt_llm::runtime::decoder"><span class="n"><span class="pre">decoder</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7decoder6OutputE" title="tensorrt_llm::runtime::decoder::Output"><span class="n"><span class="pre">Output</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">output</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7decoderE" title="tensorrt_llm::runtime::decoder"><span class="n"><span class="pre">decoder</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7decoder5InputE" title="tensorrt_llm::runtime::decoder::Input"><span class="n"><span class="pre">Input</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">input</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder7forwardERN7decoder6OutputERKN7decoder5InputE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Run one step for all requests. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime19IStatefulGptDecoder8finalizeERK14SamplingConfig">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime19IStatefulGptDecoder8finalizeERK14SamplingConfig"></span><span id="_CPPv2NK12tensorrt_llm7runtime19IStatefulGptDecoder8finalizeERK14SamplingConfig"></span><span id="tensorrt_llm::runtime::IStatefulGptDecoder::finalize__SamplingConfigCRC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IStatefulGptDecoder_1ad391989318cbc200f7b90dc0e4c6dbf2"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">finalize</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfigE" title="tensorrt_llm::runtime::SamplingConfig"><span class="n"><span class="pre">SamplingConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">samplingConfig</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime19IStatefulGptDecoder8finalizeERK14SamplingConfig" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Gather final beam search results for all requests. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime19IStatefulGptDecoder6getIdsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime19IStatefulGptDecoder6getIdsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime19IStatefulGptDecoder6getIdsEv"></span><span id="tensorrt_llm::runtime::IStatefulGptDecoder::getIdsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IStatefulGptDecoder_1aadbd74c778ac3f396021f4294dff492c"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder9TensorPtrE" title="tensorrt_llm::runtime::IStatefulGptDecoder::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getIds</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime19IStatefulGptDecoder6getIdsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[batchSize, beamWidth, maxSequenceLength], all token ids, on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime19IStatefulGptDecoder14getGatheredIdsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime19IStatefulGptDecoder14getGatheredIdsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime19IStatefulGptDecoder14getGatheredIdsEv"></span><span id="tensorrt_llm::runtime::IStatefulGptDecoder::getGatheredIdsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IStatefulGptDecoder_1a49f8c2d16936e3f2d13e58fdde4e1721"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder9TensorPtrE" title="tensorrt_llm::runtime::IStatefulGptDecoder::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getGatheredIds</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime19IStatefulGptDecoder14getGatheredIdsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[batchSize, beamWidth, maxSequenceLength] token ids after gatherTree </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime19IStatefulGptDecoder14getCumLogProbsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime19IStatefulGptDecoder14getCumLogProbsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime19IStatefulGptDecoder14getCumLogProbsEv"></span><span id="tensorrt_llm::runtime::IStatefulGptDecoder::getCumLogProbsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IStatefulGptDecoder_1ad0a22a269ec223d875f0d68bb5efc744"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder9TensorPtrE" title="tensorrt_llm::runtime::IStatefulGptDecoder::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getCumLogProbs</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime19IStatefulGptDecoder14getCumLogProbsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[batchSize, maxBeamWidth], cumulative log probabilities (per beam), on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime19IStatefulGptDecoder11getLogProbsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime19IStatefulGptDecoder11getLogProbsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime19IStatefulGptDecoder11getLogProbsEv"></span><span id="tensorrt_llm::runtime::IStatefulGptDecoder::getLogProbsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IStatefulGptDecoder_1ad1d6dd18c64d32cf98b31792c9fc499b"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder9TensorPtrE" title="tensorrt_llm::runtime::IStatefulGptDecoder::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getLogProbs</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime19IStatefulGptDecoder11getLogProbsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[batchSize, maxBeamWidth, maxSequenceLength], log probabilities (per beam), on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime19IStatefulGptDecoder12getNewTokensE10SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime19IStatefulGptDecoder12getNewTokensE10SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime19IStatefulGptDecoder12getNewTokensE10SizeType32"></span><span id="tensorrt_llm::runtime::IStatefulGptDecoder::getNewTokens__SizeType32C"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IStatefulGptDecoder_1a6871c00dca7ed1c958fa39f65afe355c"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder9TensorPtrE" title="tensorrt_llm::runtime::IStatefulGptDecoder::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getNewTokens</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">iter</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime19IStatefulGptDecoder12getNewTokensE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Get tokens generated in one step of last forward pass. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Parameters<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p><strong>iter</strong> – The iteration within [0; maxTokensPerStep) for which to get the tokens </p>
|
||
</dd>
|
||
<dt class="field-even">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-even"><p>[batchSize, beamWidth], tokens generated in <code class="docutils literal notranslate"><span class="pre">iter</span></code> (per beam), on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime19IStatefulGptDecoder15getAllNewTokensEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime19IStatefulGptDecoder15getAllNewTokensEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime19IStatefulGptDecoder15getAllNewTokensEv"></span><span id="tensorrt_llm::runtime::IStatefulGptDecoder::getAllNewTokensC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IStatefulGptDecoder_1a9ebd85b4c95c5bb2623565e4ab3983b6"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder9TensorPtrE" title="tensorrt_llm::runtime::IStatefulGptDecoder::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getAllNewTokens</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime19IStatefulGptDecoder15getAllNewTokensEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><p>Get maxTokensPerStep tokens generated in the last forward pass. </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[maxTokensPerStep, batchSize, maxBeamWidth], tokens generated in last forward pass, on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime19IStatefulGptDecoder13getNbFinishedEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime19IStatefulGptDecoder13getNbFinishedEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime19IStatefulGptDecoder13getNbFinishedEv"></span><span id="tensorrt_llm::runtime::IStatefulGptDecoder::getNbFinishedC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IStatefulGptDecoder_1a551e560fe7c9120c500f47fa84514d99"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder9TensorPtrE" title="tensorrt_llm::runtime::IStatefulGptDecoder::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getNbFinished</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime19IStatefulGptDecoder13getNbFinishedEv" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><p>[1], number of finished sequences, in pinned host memory </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoderD0Ev">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime19IStatefulGptDecoderD0Ev"></span><span id="_CPPv2N12tensorrt_llm7runtime19IStatefulGptDecoderD0Ev"></span><span id="tensorrt_llm::runtime::IStatefulGptDecoder::~IStatefulGptDecoder"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IStatefulGptDecoder_1a7dd550e0f0d980fd3f3f655f358e8ff0"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">~IStatefulGptDecoder</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">default</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoderD0Ev" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-protected-functions">Protected Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder19IStatefulGptDecoderEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime19IStatefulGptDecoder19IStatefulGptDecoderEv"></span><span id="_CPPv2N12tensorrt_llm7runtime19IStatefulGptDecoder19IStatefulGptDecoderEv"></span><span id="tensorrt_llm::runtime::IStatefulGptDecoder::IStatefulGptDecoder"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IStatefulGptDecoder_1a898fa69e15fa0d26fde3c7f50fdb8881"></span><span class="sig-name descname"><span class="n"><span class="pre">IStatefulGptDecoder</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">default</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder19IStatefulGptDecoderEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7decoderE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7decoderE"></span><span id="_CPPv2N12tensorrt_llm7runtime7decoderE"></span><span id="tensorrt_llm::runtime::decoder"></span><span class="target" id="namespacetensorrt__llm_1_1runtime_1_1decoder"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">decoder</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7decoderE" title="Link to this definition"></a><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7decoder5InputE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7decoder5InputE"></span><span id="_CPPv2N12tensorrt_llm7runtime7decoder5InputE"></span><span id="tensorrt_llm::runtime::decoder::Input"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder_1_1Input"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Input</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7decoder5InputE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7decoder5Input9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7decoder5Input9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime7decoder5Input9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder_1_1Input_1ad091c6cf3f9265c09d266c000d6ad725"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7decoder5Input9TensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7decoder5Input5InputE9TensorPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7decoder5Input5InputE9TensorPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime7decoder5Input5InputE9TensorPtr"></span><span id="tensorrt_llm::runtime::decoder::Input::Input__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder_1_1Input_1adbf9a04fb7b5c83766b236df5060d24a"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Input</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7decoder5Input9TensorPtrE" title="tensorrt_llm::runtime::decoder::Input::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">logits</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7decoder5Input5InputE9TensorPtr" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7decoder5Input6logitsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7decoder5Input6logitsE"></span><span id="_CPPv2N12tensorrt_llm7runtime7decoder5Input6logitsE"></span><span id="tensorrt_llm::runtime::decoder::Input::logits__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder_1_1Input_1a8b92e3c3e42f408503691dd5fb51e79e"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7decoder5Input9TensorPtrE" title="tensorrt_llm::runtime::decoder::Input::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">logits</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7decoder5Input6logitsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7decoder5Input16cacheIndirectionE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7decoder5Input16cacheIndirectionE"></span><span id="_CPPv2N12tensorrt_llm7runtime7decoder5Input16cacheIndirectionE"></span><span id="tensorrt_llm::runtime::decoder::Input::cacheIndirection__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder_1_1Input_1ae425ecaffbd89893ed70d3ad32c660b1"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7decoder5Input9TensorPtrE" title="tensorrt_llm::runtime::decoder::Input::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">cacheIndirection</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7decoder5Input16cacheIndirectionE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7decoder6OutputE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7decoder6OutputE"></span><span id="_CPPv2N12tensorrt_llm7runtime7decoder6OutputE"></span><span id="tensorrt_llm::runtime::decoder::Output"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder_1_1Output"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Output</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7decoder6OutputE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7decoder6Output9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7decoder6Output9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime7decoder6Output9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder_1_1Output_1a2c4dff939d33fec2fd199699c216ff28"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7decoder6Output9TensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7decoder6Output6OutputEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7decoder6Output6OutputEv"></span><span id="_CPPv2N12tensorrt_llm7runtime7decoder6Output6OutputEv"></span><span id="tensorrt_llm::runtime::decoder::Output::Output"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder_1_1Output_1af52215c47e70f6af33b6353883c24948"></span><span class="sig-name descname"><span class="n"><span class="pre">Output</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">default</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7decoder6Output6OutputEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7decoder6Output16cacheIndirectionE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7decoder6Output16cacheIndirectionE"></span><span id="_CPPv2N12tensorrt_llm7runtime7decoder6Output16cacheIndirectionE"></span><span id="tensorrt_llm::runtime::decoder::Output::cacheIndirection__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder_1_1Output_1ae9e66b08bf05167a9d16130d0518fc0e"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7decoder6Output9TensorPtrE" title="tensorrt_llm::runtime::decoder::Output::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">cacheIndirection</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7decoder6Output16cacheIndirectionE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7decoder6Output15sequenceLengthsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7decoder6Output15sequenceLengthsE"></span><span id="_CPPv2N12tensorrt_llm7runtime7decoder6Output15sequenceLengthsE"></span><span id="tensorrt_llm::runtime::decoder::Output::sequenceLengths__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder_1_1Output_1a25913f9ceca02070a650d1135ba87a72"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7decoder6Output9TensorPtrE" title="tensorrt_llm::runtime::decoder::Output::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">sequenceLengths</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7decoder6Output15sequenceLengthsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="lookaheadbuffers-h">
|
||
<h2>lookaheadBuffers.h<a class="headerlink" href="#lookaheadbuffers-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffersE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime24LookaheadDecodingBuffersE"></span><span id="_CPPv2N12tensorrt_llm7runtime24LookaheadDecodingBuffersE"></span><span id="tensorrt_llm::runtime::LookaheadDecodingBuffers"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadDecodingBuffers"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">LookaheadDecodingBuffers</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffersE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers10SizeType32E">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime24LookaheadDecodingBuffers10SizeType32E"></span><span id="_CPPv2N12tensorrt_llm7runtime24LookaheadDecodingBuffers10SizeType32E"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadDecodingBuffers_1a6b74e6912a7a384d77007bd1a32a6ecd"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">SizeType32</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers10SizeType32E" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime24LookaheadDecodingBuffers9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime24LookaheadDecodingBuffers9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadDecodingBuffers_1ae6d55e2a79a4dab469a7e846921d2e9f"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers9TensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers7ITensorE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime24LookaheadDecodingBuffers7ITensorE"></span><span id="_CPPv2N12tensorrt_llm7runtime24LookaheadDecodingBuffers7ITensorE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadDecodingBuffers_1afe60b731071b098a2e4c5097b6d467e8"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">ITensor</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers7ITensorE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers24LookaheadDecodingBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime24LookaheadDecodingBuffers24LookaheadDecodingBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerE"></span><span id="_CPPv2N12tensorrt_llm7runtime24LookaheadDecodingBuffers24LookaheadDecodingBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerE"></span><span id="tensorrt_llm::runtime::LookaheadDecodingBuffers::LookaheadDecodingBuffers__SizeType32.SizeType32.runtime::BufferManagerCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadDecodingBuffers_1a42339e7f7a0a511375d2aa7aa8cb7f77"></span><span class="sig-name descname"><span class="n"><span class="pre">LookaheadDecodingBuffers</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers10SizeType32E" title="tensorrt_llm::runtime::LookaheadDecodingBuffers::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxNumSequences</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers10SizeType32E" title="tensorrt_llm::runtime::LookaheadDecodingBuffers::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxTokensPerStep</span></span>, <a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime13BufferManagerE" title="tensorrt_llm::runtime::BufferManager"><span class="n"><span class="pre">BufferManager</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">bufferManager</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers24LookaheadDecodingBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers17generationLengthsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime24LookaheadDecodingBuffers17generationLengthsE"></span><span id="_CPPv2N12tensorrt_llm7runtime24LookaheadDecodingBuffers17generationLengthsE"></span><span id="tensorrt_llm::runtime::LookaheadDecodingBuffers::generationLengths__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadDecodingBuffers_1a386d5c3e31dd07022de404c4a54aa84a"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers9TensorPtrE" title="tensorrt_llm::runtime::LookaheadDecodingBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">generationLengths</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers17generationLengthsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers15positionOffsetsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime24LookaheadDecodingBuffers15positionOffsetsE"></span><span id="_CPPv2N12tensorrt_llm7runtime24LookaheadDecodingBuffers15positionOffsetsE"></span><span id="tensorrt_llm::runtime::LookaheadDecodingBuffers::positionOffsets__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadDecodingBuffers_1a2e45095b383e0305d81ff601a1cb7587"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers9TensorPtrE" title="tensorrt_llm::runtime::LookaheadDecodingBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">positionOffsets</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers15positionOffsetsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers11packedMasksE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime24LookaheadDecodingBuffers11packedMasksE"></span><span id="_CPPv2N12tensorrt_llm7runtime24LookaheadDecodingBuffers11packedMasksE"></span><span id="tensorrt_llm::runtime::LookaheadDecodingBuffers::packedMasks__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadDecodingBuffers_1a9ed763d83449eae9909f79dbea9b2cff"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers9TensorPtrE" title="tensorrt_llm::runtime::LookaheadDecodingBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">packedMasks</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers11packedMasksE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers11positionIdsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime24LookaheadDecodingBuffers11positionIdsE"></span><span id="_CPPv2N12tensorrt_llm7runtime24LookaheadDecodingBuffers11positionIdsE"></span><span id="tensorrt_llm::runtime::LookaheadDecodingBuffers::positionIds__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadDecodingBuffers_1aebc8b3c736dd87e008ead3c1f0e81925"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers9TensorPtrE" title="tensorrt_llm::runtime::LookaheadDecodingBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">positionIds</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffers11positionIdsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffersE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23LookaheadRuntimeBuffersE"></span><span id="_CPPv2N12tensorrt_llm7runtime23LookaheadRuntimeBuffersE"></span><span id="tensorrt_llm::runtime::LookaheadRuntimeBuffers"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadRuntimeBuffers"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">LookaheadRuntimeBuffers</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffersE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers10SizeType32E">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23LookaheadRuntimeBuffers10SizeType32E"></span><span id="_CPPv2N12tensorrt_llm7runtime23LookaheadRuntimeBuffers10SizeType32E"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadRuntimeBuffers_1ad3f7ab00996ec4856d9bf374a6c3ff4f"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">SizeType32</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers10SizeType32E" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers7ITensorE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23LookaheadRuntimeBuffers7ITensorE"></span><span id="_CPPv2N12tensorrt_llm7runtime23LookaheadRuntimeBuffers7ITensorE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadRuntimeBuffers_1a07e2c667d5f197e22f0d519671299bbf"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">ITensor</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers7ITensorE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23LookaheadRuntimeBuffers9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime23LookaheadRuntimeBuffers9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadRuntimeBuffers_1a18ac02d61fb75d3c5a2aa3ee4dbd5f3d"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorPtr</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensor9SharedPtrE" title="tensorrt_llm::runtime::ITensor::SharedPtr"><span class="n"><span class="pre">SharedPtr</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers9TensorPtrE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers9TensorMapE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23LookaheadRuntimeBuffers9TensorMapE"></span><span id="_CPPv2N12tensorrt_llm7runtime23LookaheadRuntimeBuffers9TensorMapE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadRuntimeBuffers_1a5c8b472f508f8261fccfcbbf3d4b70ec"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">TensorMap</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime12StringPtrMapE" title="tensorrt_llm::runtime::StringPtrMap"><span class="n"><span class="pre">StringPtrMap</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime7ITensorE" title="tensorrt_llm::runtime::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers9TensorMapE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers23LookaheadRuntimeBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerERKN7runtime11ModelConfigERKN7runtime11WorldConfigERKN8executor14DecodingConfigERKN7runtime11TllmRuntimeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23LookaheadRuntimeBuffers23LookaheadRuntimeBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerERKN7runtime11ModelConfigERKN7runtime11WorldConfigERKN8executor14DecodingConfigERKN7runtime11TllmRuntimeE"></span><span id="_CPPv2N12tensorrt_llm7runtime23LookaheadRuntimeBuffers23LookaheadRuntimeBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerERKN7runtime11ModelConfigERKN7runtime11WorldConfigERKN8executor14DecodingConfigERKN7runtime11TllmRuntimeE"></span><span id="tensorrt_llm::runtime::LookaheadRuntimeBuffers::LookaheadRuntimeBuffers__SizeType32.SizeType32.runtime::BufferManagerCR.runtime::ModelConfigCR.runtime::WorldConfigCR.executor::DecodingConfigCR.runtime::TllmRuntimeCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadRuntimeBuffers_1acc94c85c1404a1e3bd6cea4b403c3252"></span><span class="sig-name descname"><span class="n"><span class="pre">LookaheadRuntimeBuffers</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers10SizeType32E" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxBatchSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers10SizeType32E" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxBeamWidth</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers23LookaheadRuntimeBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerERKN7runtime11ModelConfigERKN7runtime11WorldConfigERKN8executor14DecodingConfigERKN7runtime11TllmRuntimeE" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::LookaheadRuntimeBuffers::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">BufferManager</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">manager</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers23LookaheadRuntimeBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerERKN7runtime11ModelConfigERKN7runtime11WorldConfigERKN8executor14DecodingConfigERKN7runtime11TllmRuntimeE" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::LookaheadRuntimeBuffers::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ModelConfig</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers23LookaheadRuntimeBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerERKN7runtime11ModelConfigERKN7runtime11WorldConfigERKN8executor14DecodingConfigERKN7runtime11TllmRuntimeE" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::LookaheadRuntimeBuffers::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">WorldConfig</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">worldConfig</span></span>, <a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executorE" title="tensorrt_llm::executor"><span class="n"><span class="pre">executor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executor14DecodingConfigE" title="tensorrt_llm::executor::DecodingConfig"><span class="n"><span class="pre">DecodingConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">decodingConfig</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers23LookaheadRuntimeBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerERKN7runtime11ModelConfigERKN7runtime11WorldConfigERKN8executor14DecodingConfigERKN7runtime11TllmRuntimeE" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::LookaheadRuntimeBuffers::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">TllmRuntime</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">runtime</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers23LookaheadRuntimeBuffersE10SizeType3210SizeType32RKN7runtime13BufferManagerERKN7runtime11ModelConfigERKN7runtime11WorldConfigERKN8executor14DecodingConfigERKN7runtime11TllmRuntimeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime23LookaheadRuntimeBuffers13setFromInputsE10SizeType3210SizeType32RKN7runtime7ITensorERK7ITensorRK24LookaheadDecodingBuffersRKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime23LookaheadRuntimeBuffers13setFromInputsE10SizeType3210SizeType32RKN7runtime7ITensorERK7ITensorRK24LookaheadDecodingBuffersRKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE"></span><span id="_CPPv2NK12tensorrt_llm7runtime23LookaheadRuntimeBuffers13setFromInputsE10SizeType3210SizeType32RKN7runtime7ITensorERK7ITensorRK24LookaheadDecodingBuffersRKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE"></span><span id="tensorrt_llm::runtime::LookaheadRuntimeBuffers::setFromInputs__SizeType32.SizeType32.runtime::ITensorCR.ITensorCR.LookaheadDecodingBuffersCR.runtime::TllmRuntimeCR.runtime::ModelConfigCR.runtime::WorldConfigCRC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadRuntimeBuffers_1a6efb0ff633bc62c5a5a0aedc0c6b2b41"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setFromInputs</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers10SizeType32E" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">numCtxSequences</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers10SizeType32E" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">numGenSequences</span></span>, <a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime23LookaheadRuntimeBuffers13setFromInputsE10SizeType3210SizeType32RKN7runtime7ITensorERK7ITensorRK24LookaheadDecodingBuffersRKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::setFromInputs::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ITensor</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">requestTypes</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers7ITensorE" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::ITensor"><span class="n"><span class="pre">ITensor</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">seqSlots</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime24LookaheadDecodingBuffersE" title="tensorrt_llm::runtime::LookaheadDecodingBuffers"><span class="n"><span class="pre">LookaheadDecodingBuffers</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">decoderLookaheadBuffers</span></span>, <a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime23LookaheadRuntimeBuffers13setFromInputsE10SizeType3210SizeType32RKN7runtime7ITensorERK7ITensorRK24LookaheadDecodingBuffersRKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::setFromInputs::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">TllmRuntime</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">runtime</span></span>, <a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime23LookaheadRuntimeBuffers13setFromInputsE10SizeType3210SizeType32RKN7runtime7ITensorERK7ITensorRK24LookaheadDecodingBuffersRKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::setFromInputs::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">ModelConfig</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelConfig</span></span>, <a class="reference internal" href="#_CPPv4NK12tensorrt_llm7runtime23LookaheadRuntimeBuffers13setFromInputsE10SizeType3210SizeType32RKN7runtime7ITensorERK7ITensorRK24LookaheadDecodingBuffersRKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::setFromInputs::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">WorldConfig</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">worldConfig</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime23LookaheadRuntimeBuffers13setFromInputsE10SizeType3210SizeType32RKN7runtime7ITensorERK7ITensorRK24LookaheadDecodingBuffersRKN7runtime11TllmRuntimeERKN7runtime11ModelConfigERKN7runtime11WorldConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers7reshapeE10SizeType3210SizeType3210SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23LookaheadRuntimeBuffers7reshapeE10SizeType3210SizeType3210SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime23LookaheadRuntimeBuffers7reshapeE10SizeType3210SizeType3210SizeType32"></span><span id="tensorrt_llm::runtime::LookaheadRuntimeBuffers::reshape__SizeType32.SizeType32.SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadRuntimeBuffers_1a07c7f33604a7029e91612644ad8bece5"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">reshape</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers10SizeType32E" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">numCtxSequences</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers10SizeType32E" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">numGenSequences</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers10SizeType32E" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tokensPerStep</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers7reshapeE10SizeType3210SizeType3210SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime23LookaheadRuntimeBuffers18insertInputTensorsER9TensorMapR9TensorMapRKN7runtime11WorldConfigE">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime23LookaheadRuntimeBuffers18insertInputTensorsER9TensorMapR9TensorMapRKN7runtime11WorldConfigE"></span><span id="_CPPv2NK12tensorrt_llm7runtime23LookaheadRuntimeBuffers18insertInputTensorsER9TensorMapR9TensorMapRKN7runtime11WorldConfigE"></span><span id="tensorrt_llm::runtime::LookaheadRuntimeBuffers::insertInputTensors__TensorMapR.TensorMapR.runtime::WorldConfigCRC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadRuntimeBuffers_1a392aaf845b6a6a0805a2e2b8eedcb7d2"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">insertInputTensors</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers9TensorMapE" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::TensorMap"><span class="n"><span class="pre">TensorMap</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">inputBuffers</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers9TensorMapE" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::TensorMap"><span class="n"><span class="pre">TensorMap</span></span></a><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">outputBuffers</span></span>, <a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfigE" title="tensorrt_llm::runtime::WorldConfig"><span class="n"><span class="pre">WorldConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">worldConfig</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime23LookaheadRuntimeBuffers18insertInputTensorsER9TensorMapR9TensorMapRKN7runtime11WorldConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers23enableLookaheadDecodingE10SizeType3210SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23LookaheadRuntimeBuffers23enableLookaheadDecodingE10SizeType3210SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime23LookaheadRuntimeBuffers23enableLookaheadDecodingE10SizeType3210SizeType32"></span><span id="tensorrt_llm::runtime::LookaheadRuntimeBuffers::enableLookaheadDecoding__SizeType32.SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadRuntimeBuffers_1a5b66018ff99cf49b7ac402ae11ce16ce"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">enableLookaheadDecoding</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers10SizeType32E" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxBatchSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers10SizeType32E" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tokensPerStep</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers23enableLookaheadDecodingE10SizeType3210SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers24disableLookaheadDecodingEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23LookaheadRuntimeBuffers24disableLookaheadDecodingEv"></span><span id="_CPPv2N12tensorrt_llm7runtime23LookaheadRuntimeBuffers24disableLookaheadDecodingEv"></span><span id="tensorrt_llm::runtime::LookaheadRuntimeBuffers::disableLookaheadDecoding"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadRuntimeBuffers_1a0d1b68c0a775e0629bbf3fc4d087fc96"></span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">disableLookaheadDecoding</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers24disableLookaheadDecodingEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers12cumSumLengthE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23LookaheadRuntimeBuffers12cumSumLengthE"></span><span id="_CPPv2N12tensorrt_llm7runtime23LookaheadRuntimeBuffers12cumSumLengthE"></span><span id="tensorrt_llm::runtime::LookaheadRuntimeBuffers::cumSumLength__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadRuntimeBuffers_1ac1dfeec55965185cea1df7528919f64e"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers9TensorPtrE" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">cumSumLength</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers12cumSumLengthE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers17packedMasksDeviceE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23LookaheadRuntimeBuffers17packedMasksDeviceE"></span><span id="_CPPv2N12tensorrt_llm7runtime23LookaheadRuntimeBuffers17packedMasksDeviceE"></span><span id="tensorrt_llm::runtime::LookaheadRuntimeBuffers::packedMasksDevice__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadRuntimeBuffers_1acea94f80db0e0b2a6b39b440f2ed60ec"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers9TensorPtrE" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">packedMasksDevice</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers17packedMasksDeviceE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers23generationLengthsDeviceE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23LookaheadRuntimeBuffers23generationLengthsDeviceE"></span><span id="_CPPv2N12tensorrt_llm7runtime23LookaheadRuntimeBuffers23generationLengthsDeviceE"></span><span id="tensorrt_llm::runtime::LookaheadRuntimeBuffers::generationLengthsDevice__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadRuntimeBuffers_1a09a19ccefc0db23c5c628004ac72cc1d"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers9TensorPtrE" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">generationLengthsDevice</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers23generationLengthsDeviceE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers21positionOffsetsDeviceE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23LookaheadRuntimeBuffers21positionOffsetsDeviceE"></span><span id="_CPPv2N12tensorrt_llm7runtime23LookaheadRuntimeBuffers21positionOffsetsDeviceE"></span><span id="tensorrt_llm::runtime::LookaheadRuntimeBuffers::positionOffsetsDevice__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadRuntimeBuffers_1ab984a78ad4b9b198260bcdd0141b0266"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers9TensorPtrE" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">positionOffsetsDevice</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers21positionOffsetsDeviceE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers17positionIdsDeviceE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23LookaheadRuntimeBuffers17positionIdsDeviceE"></span><span id="_CPPv2N12tensorrt_llm7runtime23LookaheadRuntimeBuffers17positionIdsDeviceE"></span><span id="tensorrt_llm::runtime::LookaheadRuntimeBuffers::positionIdsDevice__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadRuntimeBuffers_1a3d0574b682285378c72fbcc7729f1bc7"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers9TensorPtrE" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">positionIdsDevice</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers17positionIdsDeviceE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers14packedMaskHostE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23LookaheadRuntimeBuffers14packedMaskHostE"></span><span id="_CPPv2N12tensorrt_llm7runtime23LookaheadRuntimeBuffers14packedMaskHostE"></span><span id="tensorrt_llm::runtime::LookaheadRuntimeBuffers::packedMaskHost__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadRuntimeBuffers_1ab96b8e1b2a19e4899e58beb4f39d2764"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers9TensorPtrE" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">packedMaskHost</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers14packedMaskHostE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers21generationLengthsHostE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23LookaheadRuntimeBuffers21generationLengthsHostE"></span><span id="_CPPv2N12tensorrt_llm7runtime23LookaheadRuntimeBuffers21generationLengthsHostE"></span><span id="tensorrt_llm::runtime::LookaheadRuntimeBuffers::generationLengthsHost__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadRuntimeBuffers_1a01cb543b572cc39144170b48cac39266"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers9TensorPtrE" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">generationLengthsHost</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers21generationLengthsHostE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers19positionOffsetsHostE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23LookaheadRuntimeBuffers19positionOffsetsHostE"></span><span id="_CPPv2N12tensorrt_llm7runtime23LookaheadRuntimeBuffers19positionOffsetsHostE"></span><span id="tensorrt_llm::runtime::LookaheadRuntimeBuffers::positionOffsetsHost__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadRuntimeBuffers_1a768914b4e84a7f1aed192a9c7ecf99d0"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers9TensorPtrE" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">positionOffsetsHost</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers19positionOffsetsHostE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers15positionIdsHostE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23LookaheadRuntimeBuffers15positionIdsHostE"></span><span id="_CPPv2N12tensorrt_llm7runtime23LookaheadRuntimeBuffers15positionIdsHostE"></span><span id="tensorrt_llm::runtime::LookaheadRuntimeBuffers::positionIdsHost__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadRuntimeBuffers_1a30f45210b0b77d4b824249226749c8cb"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers9TensorPtrE" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">positionIdsHost</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers15positionIdsHostE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers18packedMaskHostCopyE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23LookaheadRuntimeBuffers18packedMaskHostCopyE"></span><span id="_CPPv2N12tensorrt_llm7runtime23LookaheadRuntimeBuffers18packedMaskHostCopyE"></span><span id="tensorrt_llm::runtime::LookaheadRuntimeBuffers::packedMaskHostCopy__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadRuntimeBuffers_1ae004a542179dea5feb713c4f2c2430c3"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers9TensorPtrE" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">packedMaskHostCopy</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers18packedMaskHostCopyE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers25generationLengthsHostCopyE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23LookaheadRuntimeBuffers25generationLengthsHostCopyE"></span><span id="_CPPv2N12tensorrt_llm7runtime23LookaheadRuntimeBuffers25generationLengthsHostCopyE"></span><span id="tensorrt_llm::runtime::LookaheadRuntimeBuffers::generationLengthsHostCopy__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadRuntimeBuffers_1a55fadcf67070bc31c9691f3655b0da3f"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers9TensorPtrE" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">generationLengthsHostCopy</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers25generationLengthsHostCopyE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers23positionOffsetsHostCopyE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23LookaheadRuntimeBuffers23positionOffsetsHostCopyE"></span><span id="_CPPv2N12tensorrt_llm7runtime23LookaheadRuntimeBuffers23positionOffsetsHostCopyE"></span><span id="tensorrt_llm::runtime::LookaheadRuntimeBuffers::positionOffsetsHostCopy__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadRuntimeBuffers_1a72bc7457c1183554d7796ffa8e4a1206"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers9TensorPtrE" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">positionOffsetsHostCopy</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers23positionOffsetsHostCopyE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers19positionIdsHostCopyE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23LookaheadRuntimeBuffers19positionIdsHostCopyE"></span><span id="_CPPv2N12tensorrt_llm7runtime23LookaheadRuntimeBuffers19positionIdsHostCopyE"></span><span id="tensorrt_llm::runtime::LookaheadRuntimeBuffers::positionIdsHostCopy__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadRuntimeBuffers_1ae104cf03b3a5625f3e61b99727a768b3"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers9TensorPtrE" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">positionIdsHostCopy</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers19positionIdsHostCopyE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers15useSpecDecodingE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23LookaheadRuntimeBuffers15useSpecDecodingE"></span><span id="_CPPv2N12tensorrt_llm7runtime23LookaheadRuntimeBuffers15useSpecDecodingE"></span><span id="tensorrt_llm::runtime::LookaheadRuntimeBuffers::useSpecDecoding__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadRuntimeBuffers_1a2500579fce4262a16bcbc68b77b615f9"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers9TensorPtrE" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">useSpecDecoding</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers15useSpecDecodingE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers18batchSlotsHostCopyE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime23LookaheadRuntimeBuffers18batchSlotsHostCopyE"></span><span id="_CPPv2N12tensorrt_llm7runtime23LookaheadRuntimeBuffers18batchSlotsHostCopyE"></span><span id="tensorrt_llm::runtime::LookaheadRuntimeBuffers::batchSlotsHostCopy__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadRuntimeBuffers_1a153cb9a02883c543e4779d20cfcdb72b"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers9TensorPtrE" title="tensorrt_llm::runtime::LookaheadRuntimeBuffers::TensorPtr"><span class="n"><span class="pre">TensorPtr</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">batchSlotsHostCopy</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime23LookaheadRuntimeBuffers18batchSlotsHostCopyE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="lookaheadmodule-h">
|
||
<h2>lookaheadModule.h<a class="headerlink" href="#lookaheadmodule-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15LookaheadModuleE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15LookaheadModuleE"></span><span id="_CPPv2N12tensorrt_llm7runtime15LookaheadModuleE"></span><span id="tensorrt_llm::runtime::LookaheadModule"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadModule"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">LookaheadModule</span></span></span><span class="w"> </span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="k"><span class="pre">public</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv412tensorrt_llm" title="tensorrt_llm"><span class="n"><span class="pre">tensorrt_llm</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModuleE" title="tensorrt_llm::runtime::SpeculativeDecodingModule"><span class="n"><span class="pre">SpeculativeDecodingModule</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15LookaheadModuleE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15LookaheadModule15LookaheadModuleE10SizeType3210SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15LookaheadModule15LookaheadModuleE10SizeType3210SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime15LookaheadModule15LookaheadModuleE10SizeType3210SizeType32"></span><span id="tensorrt_llm::runtime::LookaheadModule::LookaheadModule__SizeType32.SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadModule_1aa68e4d6dc548f3b5e3e218f8004be9b9"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">LookaheadModule</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxDraftPathLen</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxDecodingDraftTokens</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15LookaheadModule15LookaheadModuleE10SizeType3210SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15LookaheadModule15LookaheadModuleEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15LookaheadModule15LookaheadModuleEv"></span><span id="_CPPv2N12tensorrt_llm7runtime15LookaheadModule15LookaheadModuleEv"></span><span id="tensorrt_llm::runtime::LookaheadModule::LookaheadModule"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadModule_1a78baf71cf60b55802cf94b612644bf78"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">LookaheadModule</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15LookaheadModule15LookaheadModuleEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15LookaheadModule18setExecutionConfigERKN8executor23LookaheadDecodingConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15LookaheadModule18setExecutionConfigERKN8executor23LookaheadDecodingConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime15LookaheadModule18setExecutionConfigERKN8executor23LookaheadDecodingConfigE"></span><span id="tensorrt_llm::runtime::LookaheadModule::setExecutionConfig__executor::LookaheadDecodingConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadModule_1a38679daa2e10e8ab1f482a38a275fcff"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setExecutionConfig</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executorE" title="tensorrt_llm::executor"><span class="n"><span class="pre">executor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executor23LookaheadDecodingConfigE" title="tensorrt_llm::executor::LookaheadDecodingConfig"><span class="n"><span class="pre">LookaheadDecodingConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">config</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15LookaheadModule18setExecutionConfigERKN8executor23LookaheadDecodingConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime15LookaheadModule18getExecutionConfigEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime15LookaheadModule18getExecutionConfigEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime15LookaheadModule18getExecutionConfigEv"></span><span id="tensorrt_llm::runtime::LookaheadModule::getExecutionConfigC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadModule_1ad81b2560fd286eb36d5083279cd13f13"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executorE" title="tensorrt_llm::executor"><span class="n"><span class="pre">executor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executor23LookaheadDecodingConfigE" title="tensorrt_llm::executor::LookaheadDecodingConfig"><span class="n"><span class="pre">LookaheadDecodingConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getExecutionConfig</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime15LookaheadModule18getExecutionConfigEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-members">Private Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15LookaheadModule16mExecutionConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15LookaheadModule16mExecutionConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime15LookaheadModule16mExecutionConfigE"></span><span id="tensorrt_llm::runtime::LookaheadModule::mExecutionConfig__executor::LookaheadDecodingConfig"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1LookaheadModule_1a74f1e9093da62577f4859e740f80896e"></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executorE" title="tensorrt_llm::executor"><span class="n"><span class="pre">executor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executor23LookaheadDecodingConfigE" title="tensorrt_llm::executor::LookaheadDecodingConfig"><span class="n"><span class="pre">LookaheadDecodingConfig</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mExecutionConfig</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15LookaheadModule16mExecutionConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="modelconfig-h">
|
||
<h2>modelConfig.h<a class="headerlink" href="#modelconfig-h" title="Link to this heading"></a></h2>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfigE"></span><span id="tensorrt_llm::runtime::ModelConfig"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">ModelConfig</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-types">Public Types</p>
|
||
<dl class="cpp enum-class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig12ModelVariantE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig12ModelVariantE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig12ModelVariantE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a46a3930b326403036a2d826741aeacc7"></span><span class="k"><span class="pre">enum</span></span><span class="w"> </span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">ModelVariant</span></span></span><span class="w"> </span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">int32_t</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12ModelVariantE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p><em>Values:</em></p>
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig12ModelVariant4kGptE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig12ModelVariant4kGptE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig12ModelVariant4kGptE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a46a3930b326403036a2d826741aeacc7aeb854c1219d60212d16ad06f0d454dc8"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kGpt</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12ModelVariant4kGptE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig12ModelVariant8kChatGlmE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig12ModelVariant8kChatGlmE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig12ModelVariant8kChatGlmE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a46a3930b326403036a2d826741aeacc7a3f33cf3b2415ad1e408896306d35d786"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kChatGlm</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12ModelVariant8kChatGlmE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig12ModelVariant4kGlmE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig12ModelVariant4kGlmE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig12ModelVariant4kGlmE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a46a3930b326403036a2d826741aeacc7a30a41043cc5346b26f02b1a3e1cf54d5"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kGlm</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12ModelVariant4kGlmE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig12ModelVariant6kMambaE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig12ModelVariant6kMambaE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig12ModelVariant6kMambaE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a46a3930b326403036a2d826741aeacc7a434d259e93688631e9f679f6af07a30f"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kMamba</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12ModelVariant6kMambaE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig12ModelVariant15kRecurrentGemmaE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig12ModelVariant15kRecurrentGemmaE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig12ModelVariant15kRecurrentGemmaE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a46a3930b326403036a2d826741aeacc7a45c696d9e1b5effa650b69ac887a9983"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kRecurrentGemma</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12ModelVariant15kRecurrentGemmaE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig12ModelVariant7kEncDecE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig12ModelVariant7kEncDecE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig12ModelVariant7kEncDecE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a46a3930b326403036a2d826741aeacc7adafb6f4a6d945f658e4da3f2a8dd4fd7"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kEncDec</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12ModelVariant7kEncDecE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
<dl class="cpp enum-class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig9LayerTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig9LayerTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig9LayerTypeE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ad27648d68c7e473f9189b7843f3a8c00"></span><span class="k"><span class="pre">enum</span></span><span class="w"> </span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">LayerType</span></span></span><span class="w"> </span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">int32_t</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig9LayerTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p><em>Values:</em></p>
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig9LayerType10kATTENTIONE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig9LayerType10kATTENTIONE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig9LayerType10kATTENTIONE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ad27648d68c7e473f9189b7843f3a8c00a7c43472f56b13f7582947b4ef8e73f98"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kATTENTION</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig9LayerType10kATTENTIONE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig9LayerType10kRECURRENTE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig9LayerType10kRECURRENTE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig9LayerType10kRECURRENTE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ad27648d68c7e473f9189b7843f3a8c00a82be6fea77422f4810f7dc2cbe14bc2e"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kRECURRENT</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig9LayerType10kRECURRENTE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig9LayerType7kLINEARE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig9LayerType7kLINEARE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig9LayerType7kLINEARE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ad27648d68c7e473f9189b7843f3a8c00afd0bda8f85b35011bdcde415691fc36f"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kLINEAR</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig9LayerType7kLINEARE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig9LayerType5kNOOPE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig9LayerType5kNOOPE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig9LayerType5kNOOPE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ad27648d68c7e473f9189b7843f3a8c00ac1ccb8aad06b583f8b824ce7c4446132"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kNOOP</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig9LayerType5kNOOPE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
<dl class="cpp enum-class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig11KVCacheTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig11KVCacheTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig11KVCacheTypeE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a48d8743f0628ea2034042c6bcfed8517"></span><span class="k"><span class="pre">enum</span></span><span class="w"> </span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">KVCacheType</span></span></span><span class="w"> </span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">int32_t</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig11KVCacheTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p><em>Values:</em></p>
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig11KVCacheType11kCONTINUOUSE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig11KVCacheType11kCONTINUOUSE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig11KVCacheType11kCONTINUOUSE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a48d8743f0628ea2034042c6bcfed8517ad2fadda76e2fe15f09844dafcb790d7e"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kCONTINUOUS</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig11KVCacheType11kCONTINUOUSE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig11KVCacheType6kPAGEDE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig11KVCacheType6kPAGEDE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig11KVCacheType6kPAGEDE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a48d8743f0628ea2034042c6bcfed8517aad012df8c2260ef3b7e51d63f533dc16"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kPAGED</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig11KVCacheType6kPAGEDE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig11KVCacheType9kDISABLEDE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig11KVCacheType9kDISABLEDE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig11KVCacheType9kDISABLEDE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a48d8743f0628ea2034042c6bcfed8517a02bd32be241163867c72873cf95845a1"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kDISABLED</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig11KVCacheType9kDISABLEDE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
<dl class="cpp enum-class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig17ManageWeightsTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig17ManageWeightsTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig17ManageWeightsTypeE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1aee73725fb2f9e639356718d384025363"></span><span class="k"><span class="pre">enum</span></span><span class="w"> </span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">ManageWeightsType</span></span></span><span class="w"> </span><span class="p"><span class="pre">:</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">int32_t</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig17ManageWeightsTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd><p><em>Values:</em></p>
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig17ManageWeightsType9kDisabledE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig17ManageWeightsType9kDisabledE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig17ManageWeightsType9kDisabledE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1aee73725fb2f9e639356718d384025363a7b04d8795f5fd03c761ce70dab985fee"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kDisabled</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig17ManageWeightsType9kDisabledE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig17ManageWeightsType8kEnabledE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig17ManageWeightsType8kEnabledE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig17ManageWeightsType8kEnabledE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1aee73725fb2f9e639356718d384025363ade1ec1870756a80dffbdc4cda53153e0"></span><span class="k"><span class="pre">enumerator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kEnabled</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig17ManageWeightsType8kEnabledE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig11ModelConfigE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType3210SizeType32N8nvinfer18DataTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig11ModelConfigE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType3210SizeType32N8nvinfer18DataTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig11ModelConfigE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType3210SizeType32N8nvinfer18DataTypeE"></span><span id="tensorrt_llm::runtime::ModelConfig::ModelConfig__SizeType32.SizeType32.SizeType32.SizeType32.SizeType32.SizeType32.nvinfer1::DataType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a0beb690f207ae612e46df0f448a01225"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">ModelConfig</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">vocabSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">nbLayers</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">nbAttentionLayers</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">nbRnnLayers</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">nbHeads</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">hiddenSize</span></span>, <a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">dtype</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig11ModelConfigE10SizeType3210SizeType3210SizeType3210SizeType3210SizeType3210SizeType32N8nvinfer18DataTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig12getVocabSizeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig12getVocabSizeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig12getVocabSizeEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getVocabSizeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ad2c7526a1c003a8a4278549db295aed0"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getVocabSize</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig12getVocabSizeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig18getVocabSizePaddedE10SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig18getVocabSizePaddedE10SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig18getVocabSizePaddedE10SizeType32"></span><span id="tensorrt_llm::runtime::ModelConfig::getVocabSizePadded__SizeType32CCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a44b691295aae063983d55b0a4cf32299"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getVocabSizePadded</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">worldSize</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig18getVocabSizePaddedE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig16countLocalLayersE9LayerType10SizeType3210SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig16countLocalLayersE9LayerType10SizeType3210SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig16countLocalLayersE9LayerType10SizeType3210SizeType32"></span><span id="tensorrt_llm::runtime::ModelConfig::countLocalLayers__LayerType.SizeType32.SizeType32C"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ad422a96486029b9b47f1e346cd7d2de3"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">countLocalLayers</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig9LayerTypeE" title="tensorrt_llm::runtime::ModelConfig::LayerType"><span class="n"><span class="pre">LayerType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">layerType</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">pipelineParallelism</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">1</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">pipelineParallelismRank</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig16countLocalLayersE9LayerType10SizeType3210SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig20countLowerRankLayersE9LayerType10SizeType3210SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig20countLowerRankLayersE9LayerType10SizeType3210SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig20countLowerRankLayersE9LayerType10SizeType3210SizeType32"></span><span id="tensorrt_llm::runtime::ModelConfig::countLowerRankLayers__LayerType.SizeType32.SizeType32C"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a4c9cabd1675a0db58bce743a0ac0470e"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">countLowerRankLayers</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig9LayerTypeE" title="tensorrt_llm::runtime::ModelConfig::LayerType"><span class="n"><span class="pre">LayerType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">layerType</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">pipelineParallelism</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">1</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">pipelineParallelismRank</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig20countLowerRankLayersE9LayerType10SizeType3210SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig11getNbLayersE10SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig11getNbLayersE10SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig11getNbLayersE10SizeType32"></span><span id="tensorrt_llm::runtime::ModelConfig::getNbLayers__SizeType32C"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1aefd69a08c1409f90a4e948d857cc08b1"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getNbLayers</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">pipelineParallelism</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">1</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig11getNbLayersE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig20getNbAttentionLayersE10SizeType3210SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig20getNbAttentionLayersE10SizeType3210SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig20getNbAttentionLayersE10SizeType3210SizeType32"></span><span id="tensorrt_llm::runtime::ModelConfig::getNbAttentionLayers__SizeType32.SizeType32C"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ae7312b4e1c8f27bcba6305fa0c4eced5"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getNbAttentionLayers</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">pipelineParallelism</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">1</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">pipelineParallelismRank</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig20getNbAttentionLayersE10SizeType3210SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig14getNbRnnLayersE10SizeType3210SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig14getNbRnnLayersE10SizeType3210SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig14getNbRnnLayersE10SizeType3210SizeType32"></span><span id="tensorrt_llm::runtime::ModelConfig::getNbRnnLayers__SizeType32.SizeType32C"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ac825a4cdeeffed1ea53bec8d5fa6bba6"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getNbRnnLayers</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">pipelineParallelism</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">1</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">pipelineParallelismRank</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig14getNbRnnLayersE10SizeType3210SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig10getNbHeadsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig10getNbHeadsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig10getNbHeadsEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getNbHeadsCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ac4e85f0d390d2d281cbacc1e6d2c2a7d"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getNbHeads</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig10getNbHeadsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig12getNbKvHeadsE10SizeType32">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig12getNbKvHeadsE10SizeType32"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig12getNbKvHeadsE10SizeType32"></span><span id="tensorrt_llm::runtime::ModelConfig::getNbKvHeads__SizeType32C"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1aa5d5e18d61f0198c9c2ee02115841e81"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getNbKvHeads</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">layerIdx</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig12getNbKvHeadsE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig12setNbKvHeadsE10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig12setNbKvHeadsE10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig12setNbKvHeadsE10SizeType32"></span><span id="tensorrt_llm::runtime::ModelConfig::setNbKvHeads__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a70bc4cfb6e0ef27c086735587a275072"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setNbKvHeads</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">nbKvHeads</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12setNbKvHeadsE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig17setNbCrossKvHeadsE10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig17setNbCrossKvHeadsE10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig17setNbCrossKvHeadsE10SizeType32"></span><span id="tensorrt_llm::runtime::ModelConfig::setNbCrossKvHeads__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a89e6744eaba2b4fae56a6310dd664d2f"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setNbCrossKvHeads</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">nbKvHeads</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig17setNbCrossKvHeadsE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig13getHiddenSizeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig13getHiddenSizeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig13getHiddenSizeEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getHiddenSizeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a95aabae940f79963d6a59216cc3e8c62"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getHiddenSize</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig13getHiddenSizeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig20getEncoderHiddenSizeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig20getEncoderHiddenSizeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig20getEncoderHiddenSizeEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getEncoderHiddenSizeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1af5ff1b00564efeee5358302a34ecbb6e"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getEncoderHiddenSize</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig20getEncoderHiddenSizeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig20setEncoderHiddenSizeE10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig20setEncoderHiddenSizeE10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig20setEncoderHiddenSizeE10SizeType32"></span><span id="tensorrt_llm::runtime::ModelConfig::setEncoderHiddenSize__SizeType32CE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ae95ad1e5d827d9f266ab2c360eb91b42"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setEncoderHiddenSize</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">encoderHiddenSize</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig20setEncoderHiddenSizeE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig14getSizePerHeadEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig14getSizePerHeadEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig14getSizePerHeadEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getSizePerHeadCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1adc06556fe8bcc02ebd448a091b758cb7"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getSizePerHead</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig14getSizePerHeadEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig14setSizePerHeadE10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig14setSizePerHeadE10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig14setSizePerHeadE10SizeType32"></span><span id="tensorrt_llm::runtime::ModelConfig::setSizePerHead__SizeType32CE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a4d903cbc2da841a66e0c8f3c77bf19d7"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setSizePerHead</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">sizePerHead</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig14setSizePerHeadE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig11getDataTypeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig11getDataTypeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig11getDataTypeEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getDataTypeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ac5582fd24365d583731bbd8fc62dbe5a"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getDataType</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig11getDataTypeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig21useGptAttentionPluginEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig21useGptAttentionPluginEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig21useGptAttentionPluginEv"></span><span id="tensorrt_llm::runtime::ModelConfig::useGptAttentionPluginCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1aa4708970b05b37949e471cb815cf32b4"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">useGptAttentionPlugin</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig21useGptAttentionPluginEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig21useGptAttentionPluginEb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig21useGptAttentionPluginEb"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig21useGptAttentionPluginEb"></span><span id="tensorrt_llm::runtime::ModelConfig::useGptAttentionPlugin__bCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1af550b4ebc4dae83ea6951620d40d887d"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">useGptAttentionPlugin</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">useGptAttentionPlugin</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig21useGptAttentionPluginEb" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig20useMambaConv1dPluginEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig20useMambaConv1dPluginEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig20useMambaConv1dPluginEv"></span><span id="tensorrt_llm::runtime::ModelConfig::useMambaConv1dPluginCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1afcd536ea3183f61555d4a9bfb0a4baab"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">useMambaConv1dPlugin</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig20useMambaConv1dPluginEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig20useMambaConv1dPluginEb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig20useMambaConv1dPluginEb"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig20useMambaConv1dPluginEb"></span><span id="tensorrt_llm::runtime::ModelConfig::useMambaConv1dPlugin__bCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a6b689c7681f63d3e963680cc9f95917d"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">useMambaConv1dPlugin</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">useMambaConv1dPlugin</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig20useMambaConv1dPluginEb" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig14usePackedInputEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig14usePackedInputEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig14usePackedInputEv"></span><span id="tensorrt_llm::runtime::ModelConfig::usePackedInputCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1adafa102c76fd0db86c206f5969366038"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">usePackedInput</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig14usePackedInputEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig14usePackedInputEb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig14usePackedInputEb"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig14usePackedInputEb"></span><span id="tensorrt_llm::runtime::ModelConfig::usePackedInput__bCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a4784e5f1368098b1f78e81e4199c4985"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">usePackedInput</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">inputPacked</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig14usePackedInputEb" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig13usePagedStateEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig13usePagedStateEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig13usePagedStateEv"></span><span id="tensorrt_llm::runtime::ModelConfig::usePagedStateCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a08e0e846ca8abd00f9e929c916fcd4c0"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">usePagedState</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig13usePagedStateEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig13usePagedStateEb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig13usePagedStateEb"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig13usePagedStateEb"></span><span id="tensorrt_llm::runtime::ModelConfig::usePagedState__bCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1af20937bb5b5da5827b48174c510d15de"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">usePagedState</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">pagedState</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig13usePagedStateEb" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig17getTokensPerBlockEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig17getTokensPerBlockEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig17getTokensPerBlockEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getTokensPerBlockCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a5fc2b306388f583fbeb355a413eb5a86"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getTokensPerBlock</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig17getTokensPerBlockEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig17setTokensPerBlockE10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig17setTokensPerBlockE10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig17setTokensPerBlockE10SizeType32"></span><span id="tensorrt_llm::runtime::ModelConfig::setTokensPerBlock__SizeType32CE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a5386c582426ad3ce798605a96d889d8c"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setTokensPerBlock</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">TokensPerBlock</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig17setTokensPerBlockE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig12getQuantModeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig12getQuantModeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig12getQuantModeEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getQuantModeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1aa2ef652e4533b31ce88f4c0d1ad97121"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">common</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">QuantMode</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getQuantMode</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig12getQuantModeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig12setQuantModeEN6common9QuantModeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig12setQuantModeEN6common9QuantModeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig12setQuantModeEN6common9QuantModeE"></span><span id="tensorrt_llm::runtime::ModelConfig::setQuantMode__common::QuantModeCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ae5191acbb52d822a7dccc730af71b387"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setQuantMode</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">common</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">QuantMode</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">QuantMode</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12setQuantModeEN6common9QuantModeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig24supportsInflightBatchingEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig24supportsInflightBatchingEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig24supportsInflightBatchingEv"></span><span id="tensorrt_llm::runtime::ModelConfig::supportsInflightBatchingCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a39fb96cf15bb5175fcb35ac6158f2e06"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">supportsInflightBatching</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig24supportsInflightBatchingEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig15getMaxBatchSizeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig15getMaxBatchSizeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig15getMaxBatchSizeEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getMaxBatchSizeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1acae205bb08586324fa0ea575f349e91e"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getMaxBatchSize</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig15getMaxBatchSizeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig15setMaxBatchSizeE10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig15setMaxBatchSizeE10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig15setMaxBatchSizeE10SizeType32"></span><span id="tensorrt_llm::runtime::ModelConfig::setMaxBatchSize__SizeType32CE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1afcbe41c09cd3d1536350cc37ecc0d908"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setMaxBatchSize</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxBatchSize</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig15setMaxBatchSizeE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig15getMaxBeamWidthEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig15getMaxBeamWidthEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig15getMaxBeamWidthEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getMaxBeamWidthCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ae4eb638abf8f95b30b182b134498445c"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getMaxBeamWidth</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig15getMaxBeamWidthEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig15setMaxBeamWidthE10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig15setMaxBeamWidthE10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig15setMaxBeamWidthE10SizeType32"></span><span id="tensorrt_llm::runtime::ModelConfig::setMaxBeamWidth__SizeType32CE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a87e57088a3f8e6629d7ace77b02376d4"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setMaxBeamWidth</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxBeamWidth</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig15setMaxBeamWidthE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig14getMaxInputLenEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig14getMaxInputLenEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig14getMaxInputLenEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getMaxInputLenCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ad54640d8c03c12f153eaf02b9043463c"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getMaxInputLen</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig14getMaxInputLenEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig14setMaxInputLenE10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig14setMaxInputLenE10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig14setMaxInputLenE10SizeType32"></span><span id="tensorrt_llm::runtime::ModelConfig::setMaxInputLen__SizeType32CE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1af71d7e5e037ec2ebfe5730ed6d26feee"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setMaxInputLen</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxInputLen</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig14setMaxInputLenE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig17getMaxSequenceLenEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig17getMaxSequenceLenEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig17getMaxSequenceLenEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getMaxSequenceLenCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a8e41be40fd748e091ceb34739c21467d"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getMaxSequenceLen</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig17getMaxSequenceLenEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig17setMaxSequenceLenE10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig17setMaxSequenceLenE10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig17setMaxSequenceLenE10SizeType32"></span><span id="tensorrt_llm::runtime::ModelConfig::setMaxSequenceLen__SizeType32CE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1acf44bbab9357ab2591ffa9f218607f19"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setMaxSequenceLen</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxSequenceLen</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig17setMaxSequenceLenE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig15getMaxNumTokensEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig15getMaxNumTokensEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig15getMaxNumTokensEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getMaxNumTokensCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a837af9ec3c2c57de3fae4939029d174b"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getMaxNumTokens</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig15getMaxNumTokensEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig15setMaxNumTokensENSt8optionalI10SizeType32EE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig15setMaxNumTokensENSt8optionalI10SizeType32EE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig15setMaxNumTokensENSt8optionalI10SizeType32EE"></span><span id="tensorrt_llm::runtime::ModelConfig::setMaxNumTokens__std::optional:SizeType32:CE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1af0889ca073da922234c4edad82e71cf9"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setMaxNumTokens</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="n sig-param"><span class="pre">maxNumTokens</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig15setMaxNumTokensENSt8optionalI10SizeType32EE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig16getMaxEncoderLenEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig16getMaxEncoderLenEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig16getMaxEncoderLenEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getMaxEncoderLenCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a3c8ddbde368dce4f139b4df64da2ad06"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getMaxEncoderLen</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig16getMaxEncoderLenEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig16setMaxEncoderLenE10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig16setMaxEncoderLenE10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig16setMaxEncoderLenE10SizeType32"></span><span id="tensorrt_llm::runtime::ModelConfig::setMaxEncoderLen__SizeType32CE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a38169800b959be50c524bec9600b07c0"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setMaxEncoderLen</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxEncoderLen</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig16setMaxEncoderLenE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig15usePromptTuningEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig15usePromptTuningEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig15usePromptTuningEv"></span><span id="tensorrt_llm::runtime::ModelConfig::usePromptTuningCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a0020d543fa5a3806258f783a988ee1fc"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">usePromptTuning</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig15usePromptTuningEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig8useMropeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig8useMropeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig8useMropeEv"></span><span id="tensorrt_llm::runtime::ModelConfig::useMropeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a40cc96908df97d5a254e0c5189561ee4"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">useMrope</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig8useMropeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig11setUseMropeEb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig11setUseMropeEb"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig11setUseMropeEb"></span><span id="tensorrt_llm::runtime::ModelConfig::setUseMrope__bCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ab96860034c07d8387d3496443612e5ad"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setUseMrope</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">useMrope</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig11setUseMropeEb" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig24getMaxPositionEmbeddingsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig24getMaxPositionEmbeddingsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig24getMaxPositionEmbeddingsEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getMaxPositionEmbeddingsCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a977d19cd0ff55979f51a6d61a53c8294"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getMaxPositionEmbeddings</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig24getMaxPositionEmbeddingsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig24setMaxPositionEmbeddingsE10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig24setMaxPositionEmbeddingsE10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig24setMaxPositionEmbeddingsE10SizeType32"></span><span id="tensorrt_llm::runtime::ModelConfig::setMaxPositionEmbeddings__SizeType32CE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a3463fe9d15b9402f45d9a0477ef5ae35"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setMaxPositionEmbeddings</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxPositionEmbeddings</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig24setMaxPositionEmbeddingsE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig21getRotaryEmbeddingDimEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig21getRotaryEmbeddingDimEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig21getRotaryEmbeddingDimEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getRotaryEmbeddingDimCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a1b5f10246eed6fd56cfb04bfe8320731"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getRotaryEmbeddingDim</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig21getRotaryEmbeddingDimEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig21setRotaryEmbeddingDimE10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig21setRotaryEmbeddingDimE10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig21setRotaryEmbeddingDimE10SizeType32"></span><span id="tensorrt_llm::runtime::ModelConfig::setRotaryEmbeddingDim__SizeType32CE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ab41aa3d6b8cfecaeaeed2f4598238b49"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setRotaryEmbeddingDim</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">rotaryEmbeddingDim</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig21setRotaryEmbeddingDimE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig30getMaxPromptEmbeddingTableSizeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig30getMaxPromptEmbeddingTableSizeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig30getMaxPromptEmbeddingTableSizeEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getMaxPromptEmbeddingTableSizeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ab18461ef064f3e15265b2d823912e60c"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getMaxPromptEmbeddingTableSize</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig30getMaxPromptEmbeddingTableSizeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig30setMaxPromptEmbeddingTableSizeE10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig30setMaxPromptEmbeddingTableSizeE10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig30setMaxPromptEmbeddingTableSizeE10SizeType32"></span><span id="tensorrt_llm::runtime::ModelConfig::setMaxPromptEmbeddingTableSize__SizeType32CE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1af2fa5f70a41247541ac41becb2ccc397"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setMaxPromptEmbeddingTableSize</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxPromptEmbeddingTableSize</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig30setMaxPromptEmbeddingTableSizeE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig20computeContextLogitsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig20computeContextLogitsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig20computeContextLogitsEv"></span><span id="tensorrt_llm::runtime::ModelConfig::computeContextLogitsCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a6ff1b92b7c4079b8c27edc5916ca489c"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">computeContextLogits</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig20computeContextLogitsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig20computeContextLogitsEb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig20computeContextLogitsEb"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig20computeContextLogitsEb"></span><span id="tensorrt_llm::runtime::ModelConfig::computeContextLogits__bCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a76db846b5aa7f84e2eeb99ca291de6cd"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">computeContextLogits</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">computeContextLogits</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig20computeContextLogitsEb" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig23computeGenerationLogitsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig23computeGenerationLogitsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig23computeGenerationLogitsEv"></span><span id="tensorrt_llm::runtime::ModelConfig::computeGenerationLogitsCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1af3bad84ea1f489475d210018a5226937"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">computeGenerationLogits</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig23computeGenerationLogitsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig23computeGenerationLogitsEb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig23computeGenerationLogitsEb"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig23computeGenerationLogitsEb"></span><span id="tensorrt_llm::runtime::ModelConfig::computeGenerationLogits__bCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a81140bd79cbf1020af663df0c64f32eb"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">computeGenerationLogits</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">computeGenerationLogits</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig23computeGenerationLogitsEb" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig15getModelVariantEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig15getModelVariantEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig15getModelVariantEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getModelVariantC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1aa1c5395236e029587c0320bcdfc372da"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12ModelVariantE" title="tensorrt_llm::runtime::ModelConfig::ModelVariant"><span class="n"><span class="pre">ModelVariant</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getModelVariant</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig15getModelVariantEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig15setModelVariantE12ModelVariant">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig15setModelVariantE12ModelVariant"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig15setModelVariantE12ModelVariant"></span><span id="tensorrt_llm::runtime::ModelConfig::setModelVariant__ModelVariant"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a6b497bc99ecd2f7a41e6ccb04ee88741"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setModelVariant</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12ModelVariantE" title="tensorrt_llm::runtime::ModelConfig::ModelVariant"><span class="n"><span class="pre">ModelVariant</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">modelVariant</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig15setModelVariantE12ModelVariant" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig25getMaxDecodingDraftTokensEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig25getMaxDecodingDraftTokensEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig25getMaxDecodingDraftTokensEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getMaxDecodingDraftTokensC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a5fea542584ac51d190be4ad18ebb3489"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getMaxDecodingDraftTokens</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig25getMaxDecodingDraftTokensEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig20getMaxDecodingTokensEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig20getMaxDecodingTokensEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig20getMaxDecodingTokensEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getMaxDecodingTokensCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a3c891a1c7a4a982505fc6c6b8df196fd"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getMaxDecodingTokens</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig20getMaxDecodingTokensEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig14setContextFMHAEb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig14setContextFMHAEb"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig14setContextFMHAEb"></span><span id="tensorrt_llm::runtime::ModelConfig::setContextFMHA__bCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a462afcfe43deb94f020e90849000455e"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setContextFMHA</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">contextFMHA</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig14setContextFMHAEb" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig14getContextFMHAEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig14getContextFMHAEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig14getContextFMHAEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getContextFMHACCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a0f0cfb8b8352a61bcb47e6824f8a6103"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getContextFMHA</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig14getContextFMHAEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig19setPagedContextFMHAEb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig19setPagedContextFMHAEb"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig19setPagedContextFMHAEb"></span><span id="tensorrt_llm::runtime::ModelConfig::setPagedContextFMHA__bCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a002731c2125d9f304d4d61ba481c1ce4"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setPagedContextFMHA</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">pagedContextFMHA</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig19setPagedContextFMHAEb" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig19getPagedContextFMHAEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig19getPagedContextFMHAEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig19getPagedContextFMHAEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getPagedContextFMHACCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ae54403bdb030b67cafb6f550bed19337"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getPagedContextFMHA</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig19getPagedContextFMHAEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig18setPpReduceScatterEb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig18setPpReduceScatterEb"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig18setPpReduceScatterEb"></span><span id="tensorrt_llm::runtime::ModelConfig::setPpReduceScatter__bCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ab164555c0a5261405e87e924c59ec140"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setPpReduceScatter</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">ppReduceScatter</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig18setPpReduceScatterEb" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig18getPpReduceScatterEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig18getPpReduceScatterEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig18getPpReduceScatterEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getPpReduceScatterCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a08abc6e0cd4d9322389a9d8332ae7ea3"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getPpReduceScatter</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig18getPpReduceScatterEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig13useLoraPluginEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig13useLoraPluginEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig13useLoraPluginEv"></span><span id="tensorrt_llm::runtime::ModelConfig::useLoraPluginCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ac012addc4a1a6bbdc5c253bcd313c75a"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">useLoraPlugin</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig13useLoraPluginEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig13useLoraPluginEb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig13useLoraPluginEb"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig13useLoraPluginEb"></span><span id="tensorrt_llm::runtime::ModelConfig::useLoraPlugin__bCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1addfdf96d94cb78c38de656530bac978e"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">useLoraPlugin</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">useLoraPlugin</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig13useLoraPluginEb" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig14getLoraModulesEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig14getLoraModulesEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig14getLoraModulesEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getLoraModulesC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a394bc28cc8e8f3a925a815dcec0c8400"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModuleE" title="tensorrt_llm::runtime::LoraModule"><span class="n"><span class="pre">LoraModule</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="n"><span class="pre">getLoraModules</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig14getLoraModulesEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig14setLoraModulesERKNSt6vectorI10LoraModuleEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig14setLoraModulesERKNSt6vectorI10LoraModuleEE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig14setLoraModulesERKNSt6vectorI10LoraModuleEE"></span><span id="tensorrt_llm::runtime::ModelConfig::setLoraModules__std::vector:LoraModule:CR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1af3d088c4f0e84585ed054fa593fd4606"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setLoraModules</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModuleE" title="tensorrt_llm::runtime::LoraModule"><span class="n"><span class="pre">LoraModule</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">loraModules</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig14setLoraModulesERKNSt6vectorI10LoraModuleEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig16getMlpHiddenSizeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig16getMlpHiddenSizeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig16getMlpHiddenSizeEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getMlpHiddenSizeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a480c6e75519cd425c5fc14d7f0779139"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getMlpHiddenSize</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig16getMlpHiddenSizeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig16setMlpHiddenSizeE10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig16setMlpHiddenSizeE10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig16setMlpHiddenSizeE10SizeType32"></span><span id="tensorrt_llm::runtime::ModelConfig::setMlpHiddenSize__SizeType32CE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a45a082bcdab56ba7ffaa4179e2e8aac5"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setMlpHiddenSize</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">mlpHiddenSize</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig16setMlpHiddenSizeE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig16isKVCacheEnabledEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig16isKVCacheEnabledEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig16isKVCacheEnabledEv"></span><span id="tensorrt_llm::runtime::ModelConfig::isKVCacheEnabledCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a1c2feb8048a22aa65fa90d2fbcadd192"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">isKVCacheEnabled</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig16isKVCacheEnabledEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig14isPagedKVCacheEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig14isPagedKVCacheEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig14isPagedKVCacheEv"></span><span id="tensorrt_llm::runtime::ModelConfig::isPagedKVCacheCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a90b9ef0bd3e8cba931e27df219e945b7"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">isPagedKVCache</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig14isPagedKVCacheEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig19isContinuousKVCacheEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig19isContinuousKVCacheEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig19isContinuousKVCacheEv"></span><span id="tensorrt_llm::runtime::ModelConfig::isContinuousKVCacheCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a66964e35ccf3a678d6873fc1d67d8cfe"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">isContinuousKVCache</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig19isContinuousKVCacheEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig14getKVCacheTypeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig14getKVCacheTypeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig14getKVCacheTypeEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getKVCacheTypeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a04585fd0af4a92429771ef5ebcc831d1"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig11KVCacheTypeE" title="tensorrt_llm::runtime::ModelConfig::KVCacheType"><span class="n"><span class="pre">KVCacheType</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getKVCacheType</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig14getKVCacheTypeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig14setKVCacheTypeE11KVCacheType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig14setKVCacheTypeE11KVCacheType"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig14setKVCacheTypeE11KVCacheType"></span><span id="tensorrt_llm::runtime::ModelConfig::setKVCacheType__KVCacheTypeCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a66418c72ac9b39d9e4903c068ce3e8d9"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setKVCacheType</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig11KVCacheTypeE" title="tensorrt_llm::runtime::ModelConfig::KVCacheType"><span class="n"><span class="pre">KVCacheType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">kvCacheType</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig14setKVCacheTypeE11KVCacheType" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig17useCrossAttentionEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig17useCrossAttentionEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig17useCrossAttentionEv"></span><span id="tensorrt_llm::runtime::ModelConfig::useCrossAttentionCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1aeb95033e26e2ecabad09be0563aa07a5"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">useCrossAttention</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig17useCrossAttentionEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig20setUseCrossAttentionEb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig20setUseCrossAttentionEb"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig20setUseCrossAttentionEb"></span><span id="tensorrt_llm::runtime::ModelConfig::setUseCrossAttention__bCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a1faff3e37890eaffb65df1d29bcf15a3"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setUseCrossAttention</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">useCrossAttention</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig20setUseCrossAttentionEb" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig20usePositionEmbeddingEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig20usePositionEmbeddingEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig20usePositionEmbeddingEv"></span><span id="tensorrt_llm::runtime::ModelConfig::usePositionEmbeddingCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1adf5df806638e5486e166ce3036070060"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">usePositionEmbedding</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig20usePositionEmbeddingEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig23setUsePositionEmbeddingEb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig23setUsePositionEmbeddingEb"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig23setUsePositionEmbeddingEb"></span><span id="tensorrt_llm::runtime::ModelConfig::setUsePositionEmbedding__bCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a992b317fd23cc91af2cb8f99818fec0c"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setUsePositionEmbedding</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">usePositionEmbedding</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig23setUsePositionEmbeddingEb" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig21useTokenTypeEmbeddingEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig21useTokenTypeEmbeddingEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig21useTokenTypeEmbeddingEv"></span><span id="tensorrt_llm::runtime::ModelConfig::useTokenTypeEmbeddingCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ac8e1ec78f0746e0124ee05ec025d0509"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">useTokenTypeEmbedding</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig21useTokenTypeEmbeddingEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig24setUseTokenTypeEmbeddingEb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig24setUseTokenTypeEmbeddingEb"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig24setUseTokenTypeEmbeddingEb"></span><span id="tensorrt_llm::runtime::ModelConfig::setUseTokenTypeEmbedding__bCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a0242017a7f1a4c104e379dd1ec28213f"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setUseTokenTypeEmbedding</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">useTokenTypeEmbedding</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig24setUseTokenTypeEmbeddingEb" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig14getMaxLoraRankEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig14getMaxLoraRankEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig14getMaxLoraRankEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getMaxLoraRankCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a34ebc12fa624b6964f4aaf10ab431934"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getMaxLoraRank</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig14getMaxLoraRankEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig14setMaxLoraRankE10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig14setMaxLoraRankE10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig14setMaxLoraRankE10SizeType32"></span><span id="tensorrt_llm::runtime::ModelConfig::setMaxLoraRank__SizeType32CE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a4c2d224c8626aa07e83e9ea76dbecb7c"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setMaxLoraRank</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxLoraRank</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig14setMaxLoraRankE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig26setSpeculativeDecodingModeE23SpeculativeDecodingMode">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig26setSpeculativeDecodingModeE23SpeculativeDecodingMode"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig26setSpeculativeDecodingModeE23SpeculativeDecodingMode"></span><span id="tensorrt_llm::runtime::ModelConfig::setSpeculativeDecodingMode__SpeculativeDecodingMode"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ad7ec4cac1b5b05796701db1c18eef1c9"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setSpeculativeDecodingMode</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingModeE" title="tensorrt_llm::runtime::SpeculativeDecodingMode"><span class="n"><span class="pre">SpeculativeDecodingMode</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">mode</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig26setSpeculativeDecodingModeE23SpeculativeDecodingMode" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig28hasSpeculativeDecodingModuleEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig28hasSpeculativeDecodingModuleEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig28hasSpeculativeDecodingModuleEv"></span><span id="tensorrt_llm::runtime::ModelConfig::hasSpeculativeDecodingModuleC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a94f38f7c3078e8b5b8b965681356175c"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">hasSpeculativeDecodingModule</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig28hasSpeculativeDecodingModuleEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig28getSpeculativeDecodingModuleEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig28getSpeculativeDecodingModuleEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig28getSpeculativeDecodingModuleEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getSpeculativeDecodingModuleC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1afc7bbd0919a464d837a489771b06162f"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModuleE" title="tensorrt_llm::runtime::SpeculativeDecodingModule"><span class="n"><span class="pre">SpeculativeDecodingModule</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="n"><span class="pre">getSpeculativeDecodingModule</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig28getSpeculativeDecodingModuleEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig31getSpeculativeDecodingModulePtrEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig31getSpeculativeDecodingModulePtrEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig31getSpeculativeDecodingModulePtrEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getSpeculativeDecodingModulePtrC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a4a3cc906d7d9cca119da52b3c9609e3f"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModuleE" title="tensorrt_llm::runtime::SpeculativeDecodingModule"><span class="n"><span class="pre">SpeculativeDecodingModule</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getSpeculativeDecodingModulePtr</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig31getSpeculativeDecodingModulePtrEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig31getSpeculativeDecodingModulePtrEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig31getSpeculativeDecodingModulePtrEv"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig31getSpeculativeDecodingModulePtrEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getSpeculativeDecodingModulePtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a32cdeed2f3b2f53505371743a33f0c2d"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModuleE" title="tensorrt_llm::runtime::SpeculativeDecodingModule"><span class="n"><span class="pre">SpeculativeDecodingModule</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getSpeculativeDecodingModulePtr</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig31getSpeculativeDecodingModulePtrEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig28setSpeculativeDecodingModuleERKNSt10shared_ptrI25SpeculativeDecodingModuleEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig28setSpeculativeDecodingModuleERKNSt10shared_ptrI25SpeculativeDecodingModuleEE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig28setSpeculativeDecodingModuleERKNSt10shared_ptrI25SpeculativeDecodingModuleEE"></span><span id="tensorrt_llm::runtime::ModelConfig::setSpeculativeDecodingModule__std::shared_ptr:SpeculativeDecodingModule:CR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a9aefc7d81ff2f1e4b9ec77e16145afbe"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setSpeculativeDecodingModule</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModuleE" title="tensorrt_llm::runtime::SpeculativeDecodingModule"><span class="n"><span class="pre">SpeculativeDecodingModule</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">speculativeDecodingModule</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig28setSpeculativeDecodingModuleERKNSt10shared_ptrI25SpeculativeDecodingModuleEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig30resetSpeculativeDecodingModuleEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig30resetSpeculativeDecodingModuleEv"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig30resetSpeculativeDecodingModuleEv"></span><span id="tensorrt_llm::runtime::ModelConfig::resetSpeculativeDecodingModule"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a8fa60f44c0453831c73d10daef5da888"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">resetSpeculativeDecodingModule</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig30resetSpeculativeDecodingModuleEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig31enableSeamlessLookaheadDecodingE10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig31enableSeamlessLookaheadDecodingE10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig31enableSeamlessLookaheadDecodingE10SizeType32"></span><span id="tensorrt_llm::runtime::ModelConfig::enableSeamlessLookaheadDecoding__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a0636815cd95cc6c8fbb47f8d3195bc5a"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">enableSeamlessLookaheadDecoding</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxDraftTokens</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig31enableSeamlessLookaheadDecodingE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig32disableSeamlessLookaheadDecodingEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig32disableSeamlessLookaheadDecodingEv"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig32disableSeamlessLookaheadDecodingEv"></span><span id="tensorrt_llm::runtime::ModelConfig::disableSeamlessLookaheadDecoding"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a9de9588bdfa6b49c43448d3af1e2d27b"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">disableSeamlessLookaheadDecoding</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig32disableSeamlessLookaheadDecodingEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig13getKvDataTypeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig13getKvDataTypeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig13getKvDataTypeEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getKvDataTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a3ea1fa468f328cc71969002032deea90"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getKvDataType</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig13getKvDataTypeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig18isTransformerBasedEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig18isTransformerBasedEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig18isTransformerBasedEv"></span><span id="tensorrt_llm::runtime::ModelConfig::isTransformerBasedCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a22e6fb7c8e43e261c6d4a20b092a9e74"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">isTransformerBased</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig18isTransformerBasedEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig12hasRnnConfigEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig12hasRnnConfigEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig12hasRnnConfigEv"></span><span id="tensorrt_llm::runtime::ModelConfig::hasRnnConfigC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a25e3be93fdc314a3ee7e3e4e9eeda670"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">hasRnnConfig</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig12hasRnnConfigEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig12getRnnConfigEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig12getRnnConfigEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig12getRnnConfigEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getRnnConfigC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ac0bfbddabe94a39760e15534aef6ea00"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig9RnnConfigE" title="tensorrt_llm::runtime::ModelConfig::RnnConfig"><span class="n"><span class="pre">RnnConfig</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getRnnConfig</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig12getRnnConfigEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig12setRnnConfigERK9RnnConfig">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig12setRnnConfigERK9RnnConfig"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig12setRnnConfigERK9RnnConfig"></span><span id="tensorrt_llm::runtime::ModelConfig::setRnnConfig__RnnConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a52903902f3cac457de86a80f27a4aa40"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setRnnConfig</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig9RnnConfigE" title="tensorrt_llm::runtime::ModelConfig::RnnConfig"><span class="n"><span class="pre">RnnConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">rnnConfig</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12setRnnConfigERK9RnnConfig" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig10isRnnBasedEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig10isRnnBasedEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig10isRnnBasedEv"></span><span id="tensorrt_llm::runtime::ModelConfig::isRnnBasedCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a5e45d06e927ee447d35b7f6af188fb1d"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">isRnnBased</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig10isRnnBasedEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig13getLayerTypesEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig13getLayerTypesEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig13getLayerTypesEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getLayerTypesC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a5837dd2a49d55f5c62364a6a6601af1d"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig9LayerTypeE" title="tensorrt_llm::runtime::ModelConfig::LayerType"><span class="n"><span class="pre">LayerType</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="n"><span class="pre">getLayerTypes</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig13getLayerTypesEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig13setLayerTypesERKNSt6vectorI9LayerTypeEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig13setLayerTypesERKNSt6vectorI9LayerTypeEE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig13setLayerTypesERKNSt6vectorI9LayerTypeEE"></span><span id="tensorrt_llm::runtime::ModelConfig::setLayerTypes__std::vector:LayerType:CR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a2d5af32049283dfa801379c561dcd97f"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setLayerTypes</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig9LayerTypeE" title="tensorrt_llm::runtime::ModelConfig::LayerType"><span class="n"><span class="pre">LayerType</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">layerTypes</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig13setLayerTypesERKNSt6vectorI9LayerTypeEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig26getSpeculativeDecodingModeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig26getSpeculativeDecodingModeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig26getSpeculativeDecodingModeEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getSpeculativeDecodingModeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ab21e510daf4bdf89d432ad0fae9fb7df"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingModeE" title="tensorrt_llm::runtime::SpeculativeDecodingMode"><span class="n"><span class="pre">SpeculativeDecodingMode</span></span></a><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getSpeculativeDecodingMode</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig26getSpeculativeDecodingModeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig14setLogitsDtypeEN8nvinfer18DataTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig14setLogitsDtypeEN8nvinfer18DataTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig14setLogitsDtypeEN8nvinfer18DataTypeE"></span><span id="tensorrt_llm::runtime::ModelConfig::setLogitsDtype__nvinfer1::DataType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a521abb493b4b844a44ccd94121d2b9ed"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setLogitsDtype</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">inputDtype</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig14setLogitsDtypeEN8nvinfer18DataTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig14getLogitsDtypeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig14getLogitsDtypeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig14getLogitsDtypeEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getLogitsDtypeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1adbf3366f2f80450af6a42bfa5f2942f7"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getLogitsDtype</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig14getLogitsDtypeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig20setUseShapeInferenceEb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig20setUseShapeInferenceEb"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig20setUseShapeInferenceEb"></span><span id="tensorrt_llm::runtime::ModelConfig::setUseShapeInference__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1af9020a3352a95c41af6ccf856fc63740"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setUseShapeInference</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">useShapeInference</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig20setUseShapeInferenceEb" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig17useShapeInferenceEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig17useShapeInferenceEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig17useShapeInferenceEv"></span><span id="tensorrt_llm::runtime::ModelConfig::useShapeInferenceC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ac3900be3c0184e7674b6ca9cf1c1e24f"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">useShapeInference</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig17useShapeInferenceEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig20getManageWeightsTypeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig20getManageWeightsTypeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig20getManageWeightsTypeEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getManageWeightsTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a5ea16427c0ef9dc4e4cb3ab297400684"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig17ManageWeightsTypeE" title="tensorrt_llm::runtime::ModelConfig::ManageWeightsType"><span class="n"><span class="pre">ManageWeightsType</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getManageWeightsType</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig20getManageWeightsTypeEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig20setManageWeightsTypeEK17ManageWeightsType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig20setManageWeightsTypeEK17ManageWeightsType"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig20setManageWeightsTypeEK17ManageWeightsType"></span><span id="tensorrt_llm::runtime::ModelConfig::setManageWeightsType__ManageWeightsTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ac8aee3aa46d172b28e9e61cc177f5c57"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setManageWeightsType</span></span></span><span class="sig-paren">(</span><span class="k"><span class="pre">const</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig17ManageWeightsTypeE" title="tensorrt_llm::runtime::ModelConfig::ManageWeightsType"><span class="n"><span class="pre">ManageWeightsType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">manageWeightType</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig20setManageWeightsTypeEK17ManageWeightsType" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig12getModelNameEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig12getModelNameEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig12getModelNameEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getModelNameC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a38ff64aeb32deaa59009048788ef7205"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="n"><span class="pre">getModelName</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig12getModelNameEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig12setModelNameERKNSt6stringE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig12setModelNameERKNSt6stringE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig12setModelNameERKNSt6stringE"></span><span id="tensorrt_llm::runtime::ModelConfig::setModelName__ssCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a466ea60c644630d431b4db7882123756"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setModelName</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">modelName</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12setModelNameERKNSt6stringE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig21getNumKvHeadsPerLayerEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig21getNumKvHeadsPerLayerEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig21getNumKvHeadsPerLayerEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getNumKvHeadsPerLayerC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a2d5f312460c7acec54548c8530342f3a"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="n"><span class="pre">getNumKvHeadsPerLayer</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig21getNumKvHeadsPerLayerEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig31getNumKvHeadsPerLayerLocalRangeE10SizeType3210SizeType32b">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig31getNumKvHeadsPerLayerLocalRangeE10SizeType3210SizeType32b"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig31getNumKvHeadsPerLayerLocalRangeE10SizeType3210SizeType32b"></span><span id="tensorrt_llm::runtime::ModelConfig::getNumKvHeadsPerLayerLocalRange__SizeType32.SizeType32.bC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a05ea49becb87faba3755e95a98484aac"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">pair</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">const_iterator</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">const_iterator</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getNumKvHeadsPerLayerLocalRange</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">pipelineParallelism</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">1</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">pipelineParallelismRank</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span>, <span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">isCrossAttention</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">false</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig31getNumKvHeadsPerLayerLocalRangeE10SizeType3210SizeType32b" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig21setNumKvHeadsPerLayerERKNSt6vectorI10SizeType32EE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig21setNumKvHeadsPerLayerERKNSt6vectorI10SizeType32EE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig21setNumKvHeadsPerLayerERKNSt6vectorI10SizeType32EE"></span><span id="tensorrt_llm::runtime::ModelConfig::setNumKvHeadsPerLayer__std::vector:SizeType32:CR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ad0711ba3755ba943f52a8bfee4adffcb"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setNumKvHeadsPerLayer</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">headsPerLayer</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig21setNumKvHeadsPerLayerERKNSt6vectorI10SizeType32EE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig26setNumKvHeadsPerCrossLayerERKNSt6vectorI10SizeType32EE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig26setNumKvHeadsPerCrossLayerERKNSt6vectorI10SizeType32EE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig26setNumKvHeadsPerCrossLayerERKNSt6vectorI10SizeType32EE"></span><span id="tensorrt_llm::runtime::ModelConfig::setNumKvHeadsPerCrossLayer__std::vector:SizeType32:CR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a59b7dba50c19d9e418a9816b668340e4"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setNumKvHeadsPerCrossLayer</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">headsPerLayer</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig26setNumKvHeadsPerCrossLayerERKNSt6vectorI10SizeType32EE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig18getSumLocalKvHeadsE10SizeType3210SizeType32b">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig18getSumLocalKvHeadsE10SizeType3210SizeType32b"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig18getSumLocalKvHeadsE10SizeType3210SizeType32b"></span><span id="tensorrt_llm::runtime::ModelConfig::getSumLocalKvHeads__SizeType32.SizeType32.bC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a639038d8052a0789c54b0d6c7f1ee040"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getSumLocalKvHeads</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">pipelineParallelism</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">1</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">pipelineParallelismRank</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span>, <span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">isCrossAttention</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="k"><span class="pre">false</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig18getSumLocalKvHeadsE10SizeType3210SizeType32b" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11ModelConfig19skipCrossAttnBlocksEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11ModelConfig19skipCrossAttnBlocksEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11ModelConfig19skipCrossAttnBlocksEv"></span><span id="tensorrt_llm::runtime::ModelConfig::skipCrossAttnBlocksCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a144f4f5b81ad7d6d42159033418f1a94"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">skipCrossAttnBlocks</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime11ModelConfig19skipCrossAttnBlocksEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig22setSkipCrossAttnBlocksEb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig22setSkipCrossAttnBlocksEb"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig22setSkipCrossAttnBlocksEb"></span><span id="tensorrt_llm::runtime::ModelConfig::setSkipCrossAttnBlocks__bCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a44877e23b091833fcc848a92be8e53f1"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">void</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">setSkipCrossAttnBlocks</span></span></span><span class="sig-paren">(</span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">skipCrossAttnBlocks</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig22setSkipCrossAttnBlocksEb" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-functions">Public Static Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig21KVCacheTypeFromStringENSt6stringE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig21KVCacheTypeFromStringENSt6stringE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig21KVCacheTypeFromStringENSt6stringE"></span><span id="tensorrt_llm::runtime::ModelConfig::KVCacheTypeFromString__ss"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1adc0f27d15df1f1467da047c129056d0b"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig11KVCacheTypeE" title="tensorrt_llm::runtime::ModelConfig::KVCacheType"><span class="n"><span class="pre">KVCacheType</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">KVCacheTypeFromString</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">value</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig21KVCacheTypeFromStringENSt6stringE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig25getOptProfilesSplitPointsEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig25getOptProfilesSplitPointsEv"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig25getOptProfilesSplitPointsEv"></span><span id="tensorrt_llm::runtime::ModelConfig::getOptProfilesSplitPoints"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a63b9edc5318d4fbd754446da66ec93dc"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getOptProfilesSplitPoints</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">noexcept</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig25getOptProfilesSplitPointsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-static-attributes">Public Static Attributes</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig26kOPT_PROFILES_SPLIT_POINTSE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig26kOPT_PROFILES_SPLIT_POINTSE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig26kOPT_PROFILES_SPLIT_POINTSE"></span><span id="tensorrt_llm::runtime::ModelConfig::kOPT_PROFILES_SPLIT_POINTS__std::array"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a88507c9a7c6f52dc95eec39741eb8d47"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">array</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kOPT_PROFILES_SPLIT_POINTS</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="m"><span class="pre">64</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">128</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">256</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">512</span></span><span class="p"><span class="pre">,</span></span><span class="w"> </span><span class="m"><span class="pre">1024</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig26kOPT_PROFILES_SPLIT_POINTSE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig29kDEFAULT_NUM_TOKENS_PER_BLOCKE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig29kDEFAULT_NUM_TOKENS_PER_BLOCKE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig29kDEFAULT_NUM_TOKENS_PER_BLOCKE"></span><span id="tensorrt_llm::runtime::ModelConfig::kDEFAULT_NUM_TOKENS_PER_BLOCK__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a206f572f528fa3e31abc71080f1aa2cd"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">constexpr</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">kDEFAULT_NUM_TOKENS_PER_BLOCK</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">64</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig29kDEFAULT_NUM_TOKENS_PER_BLOCKE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-members">Private Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig10mVocabSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig10mVocabSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig10mVocabSizeE"></span><span id="tensorrt_llm::runtime::ModelConfig::mVocabSize__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a37a3d0383c8a5d366a6d914be237b971"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mVocabSize</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig10mVocabSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig9mNbLayersE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig9mNbLayersE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig9mNbLayersE"></span><span id="tensorrt_llm::runtime::ModelConfig::mNbLayers__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a86f9c01d713df2e5ac6dbf5c96e256ca"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mNbLayers</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig9mNbLayersE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig18mNbAttentionLayersE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig18mNbAttentionLayersE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig18mNbAttentionLayersE"></span><span id="tensorrt_llm::runtime::ModelConfig::mNbAttentionLayers__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1aff01f3f88389157b57553c5d73d3f8c6"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mNbAttentionLayers</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig18mNbAttentionLayersE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig12mNbRnnLayersE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig12mNbRnnLayersE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig12mNbRnnLayersE"></span><span id="tensorrt_llm::runtime::ModelConfig::mNbRnnLayers__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a61c3f51a17a377b956834b6b716fb606"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mNbRnnLayers</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12mNbRnnLayersE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig8mNbHeadsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig8mNbHeadsE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig8mNbHeadsE"></span><span id="tensorrt_llm::runtime::ModelConfig::mNbHeads__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a2c02a45e1219d6b00dbcf3637fcaf231"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mNbHeads</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig8mNbHeadsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig11mHiddenSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig11mHiddenSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig11mHiddenSizeE"></span><span id="tensorrt_llm::runtime::ModelConfig::mHiddenSize__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a2121e3fbad27dd5672a58c6214accaa0"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mHiddenSize</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig11mHiddenSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig12mSizePerHeadE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig12mSizePerHeadE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig12mSizePerHeadE"></span><span id="tensorrt_llm::runtime::ModelConfig::mSizePerHead__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a21af1497012f9bddd323992c1a04d677"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mSizePerHead</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12mSizePerHeadE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig9mDataTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig9mDataTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig9mDataTypeE"></span><span id="tensorrt_llm::runtime::ModelConfig::mDataType__nvinfer1::DataType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1affb02d8d91bd93ae9ccb486bbdb2abfe"></span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mDataType</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig9mDataTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig22mUseGptAttentionPluginE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig22mUseGptAttentionPluginE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig22mUseGptAttentionPluginE"></span><span id="tensorrt_llm::runtime::ModelConfig::mUseGptAttentionPlugin__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1aad9e2f0b2d194a09bdb97fad5ea34cf8"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mUseGptAttentionPlugin</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig22mUseGptAttentionPluginE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig21mUseMambaConv1dPluginE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig21mUseMambaConv1dPluginE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig21mUseMambaConv1dPluginE"></span><span id="tensorrt_llm::runtime::ModelConfig::mUseMambaConv1dPlugin__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a36680ffd87ff6fcb1d04850d5e8bf7bf"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mUseMambaConv1dPlugin</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig21mUseMambaConv1dPluginE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig12mInputPackedE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig12mInputPackedE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig12mInputPackedE"></span><span id="tensorrt_llm::runtime::ModelConfig::mInputPacked__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a882a61ec0abd6731fe39e89b3842f22e"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mInputPacked</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12mInputPackedE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig11mPagedStateE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig11mPagedStateE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig11mPagedStateE"></span><span id="tensorrt_llm::runtime::ModelConfig::mPagedState__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a8ecdf08d955ce84fb86d1a1b2e2ccf08"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mPagedState</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig11mPagedStateE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig15mTokensPerBlockE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig15mTokensPerBlockE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig15mTokensPerBlockE"></span><span id="tensorrt_llm::runtime::ModelConfig::mTokensPerBlock__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1aa046f0031177bbcb63ce5897ebcddde2"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mTokensPerBlock</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig15mTokensPerBlockE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig10mQuantModeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig10mQuantModeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig10mQuantModeE"></span><span id="tensorrt_llm::runtime::ModelConfig::mQuantMode__common::QuantMode"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a61b5f8cdf8fccb1c130d260033e2ff00"></span><span class="n"><span class="pre">common</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">QuantMode</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mQuantMode</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig10mQuantModeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig13mMaxBatchSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig13mMaxBatchSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig13mMaxBatchSizeE"></span><span id="tensorrt_llm::runtime::ModelConfig::mMaxBatchSize__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a61a5e9d7334719dd85876b36bb6b15a2"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mMaxBatchSize</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig13mMaxBatchSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig13mMaxBeamWidthE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig13mMaxBeamWidthE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig13mMaxBeamWidthE"></span><span id="tensorrt_llm::runtime::ModelConfig::mMaxBeamWidth__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a1813511a528d48f49e9e8e3aa3c0352c"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mMaxBeamWidth</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig13mMaxBeamWidthE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig12mMaxInputLenE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig12mMaxInputLenE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig12mMaxInputLenE"></span><span id="tensorrt_llm::runtime::ModelConfig::mMaxInputLen__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1aa93bc77b0e2f29dc570da55dc23e5da3"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mMaxInputLen</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12mMaxInputLenE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig15mMaxSequenceLenE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig15mMaxSequenceLenE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig15mMaxSequenceLenE"></span><span id="tensorrt_llm::runtime::ModelConfig::mMaxSequenceLen__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1aa5aceb745909848d3a44e6d7d25e1154"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mMaxSequenceLen</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig15mMaxSequenceLenE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig13mMaxNumTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig13mMaxNumTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig13mMaxNumTokensE"></span><span id="tensorrt_llm::runtime::ModelConfig::mMaxNumTokens__std::optional:SizeType32:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a28f632083f25f9916316be1ea4b7376b"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mMaxNumTokens</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig13mMaxNumTokensE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig21mComputeContextLogitsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig21mComputeContextLogitsE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig21mComputeContextLogitsE"></span><span id="tensorrt_llm::runtime::ModelConfig::mComputeContextLogits__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a62f06f9d9223d6ce86537666aabab262"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mComputeContextLogits</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig21mComputeContextLogitsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig24mComputeGenerationLogitsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig24mComputeGenerationLogitsE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig24mComputeGenerationLogitsE"></span><span id="tensorrt_llm::runtime::ModelConfig::mComputeGenerationLogits__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a8fc084e75b3e7163e1a55431e911e71f"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mComputeGenerationLogits</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig24mComputeGenerationLogitsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig13mModelVariantE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig13mModelVariantE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig13mModelVariantE"></span><span id="tensorrt_llm::runtime::ModelConfig::mModelVariant__ModelVariant"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ad1b58680292bdf1913810e12f031fd09"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12ModelVariantE" title="tensorrt_llm::runtime::ModelConfig::ModelVariant"><span class="n"><span class="pre">ModelVariant</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mModelVariant</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig13mModelVariantE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig28mMaxPromptEmbeddingTableSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig28mMaxPromptEmbeddingTableSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig28mMaxPromptEmbeddingTableSizeE"></span><span id="tensorrt_llm::runtime::ModelConfig::mMaxPromptEmbeddingTableSize__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ad4bf7c9661e68d3b2711d90ff943fe83"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mMaxPromptEmbeddingTableSize</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig28mMaxPromptEmbeddingTableSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig9mUseMropeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig9mUseMropeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig9mUseMropeE"></span><span id="tensorrt_llm::runtime::ModelConfig::mUseMrope__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a532a93789582c93ce8da37d7672904a5"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mUseMrope</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig9mUseMropeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig22mMaxPositionEmbeddingsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig22mMaxPositionEmbeddingsE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig22mMaxPositionEmbeddingsE"></span><span id="tensorrt_llm::runtime::ModelConfig::mMaxPositionEmbeddings__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1abb4f2bc166989ca1ddfc7ae5485f7386"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mMaxPositionEmbeddings</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig22mMaxPositionEmbeddingsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig19mRotaryEmbeddingDimE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig19mRotaryEmbeddingDimE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig19mRotaryEmbeddingDimE"></span><span id="tensorrt_llm::runtime::ModelConfig::mRotaryEmbeddingDim__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a8f2487f2464369a6e503fefcf92d34b5"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mRotaryEmbeddingDim</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig19mRotaryEmbeddingDimE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig12mContextFMHAE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig12mContextFMHAE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig12mContextFMHAE"></span><span id="tensorrt_llm::runtime::ModelConfig::mContextFMHA__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a9571d6946c7b8a64d6e69bf6746ff639"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mContextFMHA</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12mContextFMHAE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig17mPagedContextFMHAE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig17mPagedContextFMHAE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig17mPagedContextFMHAE"></span><span id="tensorrt_llm::runtime::ModelConfig::mPagedContextFMHA__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a3f6ec1842133c50530dd26c9faaf7a52"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mPagedContextFMHA</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig17mPagedContextFMHAE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig16mPpReduceScatterE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig16mPpReduceScatterE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig16mPpReduceScatterE"></span><span id="tensorrt_llm::runtime::ModelConfig::mPpReduceScatter__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ab258eb7446c0efa12a53d655e97ec86c"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mPpReduceScatter</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig16mPpReduceScatterE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig14mUseLoraPluginE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig14mUseLoraPluginE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig14mUseLoraPluginE"></span><span id="tensorrt_llm::runtime::ModelConfig::mUseLoraPlugin__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a3be94f3bdf5c5e2e08482f98537d3d90"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mUseLoraPlugin</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig14mUseLoraPluginE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig12mLoraModulesE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig12mLoraModulesE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig12mLoraModulesE"></span><span id="tensorrt_llm::runtime::ModelConfig::mLoraModules__std::vector:LoraModule:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a34b85b164e6d90afaa45ff133b6c8c03"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10LoraModuleE" title="tensorrt_llm::runtime::LoraModule"><span class="n"><span class="pre">LoraModule</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mLoraModules</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12mLoraModulesE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig14mMlpHiddenSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig14mMlpHiddenSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig14mMlpHiddenSizeE"></span><span id="tensorrt_llm::runtime::ModelConfig::mMlpHiddenSize__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a54508516a416b2ae44d1d82253ad9678"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mMlpHiddenSize</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig14mMlpHiddenSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig12mMaxLoraRankE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig12mMaxLoraRankE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig12mMaxLoraRankE"></span><span id="tensorrt_llm::runtime::ModelConfig::mMaxLoraRank__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a78d71eb5426538c3e343cbc92fbbdabe"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mMaxLoraRank</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12mMaxLoraRankE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig10mRnnConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig10mRnnConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig10mRnnConfigE"></span><span id="tensorrt_llm::runtime::ModelConfig::mRnnConfig__std::optional:RnnConfig:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a2376ed37b3e21ced6c540e83aba6ca60"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig9RnnConfigE" title="tensorrt_llm::runtime::ModelConfig::RnnConfig"><span class="n"><span class="pre">RnnConfig</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mRnnConfig</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig10mRnnConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig12mKVCacheTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig12mKVCacheTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig12mKVCacheTypeE"></span><span id="tensorrt_llm::runtime::ModelConfig::mKVCacheType__KVCacheType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a3ec8c245cfa56077d52e97fe55d6e75d"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig11KVCacheTypeE" title="tensorrt_llm::runtime::ModelConfig::KVCacheType"><span class="n"><span class="pre">KVCacheType</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mKVCacheType</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig11KVCacheTypeE" title="tensorrt_llm::runtime::ModelConfig::KVCacheType"><span class="n"><span class="pre">KVCacheType</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig11KVCacheType11kCONTINUOUSE" title="tensorrt_llm::runtime::ModelConfig::KVCacheType::kCONTINUOUS"><span class="n"><span class="pre">kCONTINUOUS</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12mKVCacheTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig14mMaxEncoderLenE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig14mMaxEncoderLenE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig14mMaxEncoderLenE"></span><span id="tensorrt_llm::runtime::ModelConfig::mMaxEncoderLen__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a8c4cdc68888b85fbd943e5edeb7922c7"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mMaxEncoderLen</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig14mMaxEncoderLenE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig18mEncoderHiddenSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig18mEncoderHiddenSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig18mEncoderHiddenSizeE"></span><span id="tensorrt_llm::runtime::ModelConfig::mEncoderHiddenSize__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a431ae6b9afe40b5bd4d2206a5cc7840d"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mEncoderHiddenSize</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="p"><span class="pre">{</span></span><span class="p"><span class="pre">}</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig18mEncoderHiddenSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig18mUseCrossAttentionE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig18mUseCrossAttentionE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig18mUseCrossAttentionE"></span><span id="tensorrt_llm::runtime::ModelConfig::mUseCrossAttention__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1abf9500bc966e156bf0d476836d2b572a"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mUseCrossAttention</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig18mUseCrossAttentionE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig21mUsePositionEmbeddingE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig21mUsePositionEmbeddingE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig21mUsePositionEmbeddingE"></span><span id="tensorrt_llm::runtime::ModelConfig::mUsePositionEmbedding__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ac1492ed33343b577717489050a8a4b5c"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mUsePositionEmbedding</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig21mUsePositionEmbeddingE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig22mUseTokenTypeEmbeddingE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig22mUseTokenTypeEmbeddingE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig22mUseTokenTypeEmbeddingE"></span><span id="tensorrt_llm::runtime::ModelConfig::mUseTokenTypeEmbedding__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1af6dc99e292a0518d5fcd1e221786d603"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mUseTokenTypeEmbedding</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig22mUseTokenTypeEmbeddingE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig11mLayerTypesE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig11mLayerTypesE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig11mLayerTypesE"></span><span id="tensorrt_llm::runtime::ModelConfig::mLayerTypes__std::vector:LayerType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a147a322e201414cb6fc34d0ed8706e68"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig9LayerTypeE" title="tensorrt_llm::runtime::ModelConfig::LayerType"><span class="n"><span class="pre">LayerType</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mLayerTypes</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig11mLayerTypesE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig26mSpeculativeDecodingModuleE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig26mSpeculativeDecodingModuleE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig26mSpeculativeDecodingModuleE"></span><span id="tensorrt_llm::runtime::ModelConfig::mSpeculativeDecodingModule__std::shared_ptr:SpeculativeDecodingModule:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a6835321ce94e9fd421c391f1b5130a61"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">shared_ptr</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime25SpeculativeDecodingModuleE" title="tensorrt_llm::runtime::SpeculativeDecodingModule"><span class="n"><span class="pre">SpeculativeDecodingModule</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mSpeculativeDecodingModule</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig26mSpeculativeDecodingModuleE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig24mSpeculativeDecodingModeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig24mSpeculativeDecodingModeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig24mSpeculativeDecodingModeE"></span><span id="tensorrt_llm::runtime::ModelConfig::mSpeculativeDecodingMode__SpeculativeDecodingMode"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a8972453f4a38857ff67333c8736b36a3"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime23SpeculativeDecodingModeE" title="tensorrt_llm::runtime::SpeculativeDecodingMode"><span class="n"><span class="pre">SpeculativeDecodingMode</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mSpeculativeDecodingMode</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig24mSpeculativeDecodingModeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig12mLogitsDtypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig12mLogitsDtypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig12mLogitsDtypeE"></span><span id="tensorrt_llm::runtime::ModelConfig::mLogitsDtype__nvinfer1::DataType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a031d0a099cfde1aea2f1181dd135f084"></span><a class="reference internal" href="#_CPPv48nvinfer1" title="nvinfer1"><span class="n"><span class="pre">nvinfer1</span></span></a><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">DataType</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mLogitsDtype</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig12mLogitsDtypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig18mUseShapeInferenceE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig18mUseShapeInferenceE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig18mUseShapeInferenceE"></span><span id="tensorrt_llm::runtime::ModelConfig::mUseShapeInference__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1ae913a073cfb1a11c28886cb567c5900b"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mUseShapeInference</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig18mUseShapeInferenceE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig18mManageWeightsTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig18mManageWeightsTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig18mManageWeightsTypeE"></span><span id="tensorrt_llm::runtime::ModelConfig::mManageWeightsType__ManageWeightsType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a64d9addbd71e0ed7d43db7b035ed5c11"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig17ManageWeightsTypeE" title="tensorrt_llm::runtime::ModelConfig::ManageWeightsType"><span class="n"><span class="pre">ManageWeightsType</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mManageWeightsType</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig18mManageWeightsTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig10mModelNameE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig10mModelNameE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig10mModelNameE"></span><span id="tensorrt_llm::runtime::ModelConfig::mModelName__ss"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a040336df17e786e007f91dac8d0a5f4c"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mModelName</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig10mModelNameE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig28mNumKvHeadsPerAttentionLayerE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig28mNumKvHeadsPerAttentionLayerE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig28mNumKvHeadsPerAttentionLayerE"></span><span id="tensorrt_llm::runtime::ModelConfig::mNumKvHeadsPerAttentionLayer__std::vector:SizeType32:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a476d2ac4959e8974ea5779c9b66387a3"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mNumKvHeadsPerAttentionLayer</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig28mNumKvHeadsPerAttentionLayerE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig33mNumKvHeadsPerCrossAttentionLayerE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig33mNumKvHeadsPerCrossAttentionLayerE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig33mNumKvHeadsPerCrossAttentionLayerE"></span><span id="tensorrt_llm::runtime::ModelConfig::mNumKvHeadsPerCrossAttentionLayer__std::vector:SizeType32:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a6295c0b948ef1c4220c1248a7f54b9f2"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mNumKvHeadsPerCrossAttentionLayer</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig33mNumKvHeadsPerCrossAttentionLayerE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig20mSkipCrossAttnBlocksE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig20mSkipCrossAttnBlocksE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig20mSkipCrossAttnBlocksE"></span><span id="tensorrt_llm::runtime::ModelConfig::mSkipCrossAttnBlocks__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ModelConfig_1a4201343e4746e9f733c5b82c0a395aa3"></span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mSkipCrossAttnBlocks</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig20mSkipCrossAttnBlocksE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig9RnnConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig9RnnConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig9RnnConfigE"></span><span id="tensorrt_llm::runtime::ModelConfig::RnnConfig"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1ModelConfig_1_1RnnConfig"></span><span class="k"><span class="pre">struct</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">RnnConfig</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig9RnnConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig9RnnConfig9stateSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig9RnnConfig9stateSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig9RnnConfig9stateSizeE"></span><span id="tensorrt_llm::runtime::ModelConfig::RnnConfig::stateSize__SizeType32"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1ModelConfig_1_1RnnConfig_1aa48db3c2a457acac081d0464eb810d10"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">stateSize</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig9RnnConfig9stateSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig9RnnConfig10convKernelE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig9RnnConfig10convKernelE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig9RnnConfig10convKernelE"></span><span id="tensorrt_llm::runtime::ModelConfig::RnnConfig::convKernel__SizeType32"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1ModelConfig_1_1RnnConfig_1a95f342da763c9eccf41f82022d6aaece"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">convKernel</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig9RnnConfig10convKernelE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig9RnnConfig13rnnHiddenSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig9RnnConfig13rnnHiddenSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig9RnnConfig13rnnHiddenSizeE"></span><span id="tensorrt_llm::runtime::ModelConfig::RnnConfig::rnnHiddenSize__SizeType32"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1ModelConfig_1_1RnnConfig_1a563fdfb3179e823cebf6a29bfa79af6f"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">rnnHiddenSize</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig9RnnConfig13rnnHiddenSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig9RnnConfig11rnnHeadSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig9RnnConfig11rnnHeadSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig9RnnConfig11rnnHeadSizeE"></span><span id="tensorrt_llm::runtime::ModelConfig::RnnConfig::rnnHeadSize__SizeType32"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1ModelConfig_1_1RnnConfig_1a224165c5e8b9a336c97c0630ef87b5af"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">rnnHeadSize</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig9RnnConfig11rnnHeadSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11ModelConfig9RnnConfig14rnnConvDimSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11ModelConfig9RnnConfig14rnnConvDimSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11ModelConfig9RnnConfig14rnnConvDimSizeE"></span><span id="tensorrt_llm::runtime::ModelConfig::RnnConfig::rnnConvDimSize__SizeType32"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1ModelConfig_1_1RnnConfig_1af3e1e4f57fd7ce7585b9d6466125c6dd"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">rnnConvDimSize</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11ModelConfig9RnnConfig14rnnConvDimSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="samplingconfig-h">
|
||
<h2>samplingConfig.h<a class="headerlink" href="#samplingconfig-h" title="Link to this heading"></a></h2>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-defines">Defines</p>
|
||
<dl class="cpp macro">
|
||
<dt class="sig sig-object cpp" id="c.SET_FROM_OPTIONAL">
|
||
<span class="target" id="samplingConfig_8h_1ae2a9ee9b68dd5ae5302af8d853bab8da"></span><span class="sig-name descname"><span class="n"><span class="pre">SET_FROM_OPTIONAL</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">varName</span></span>, <span class="n"><span class="pre">VarName</span></span>, <span class="n"><span class="pre">VarType</span></span><span class="sig-paren">)</span><a class="headerlink" href="#c.SET_FROM_OPTIONAL" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">tensorrt_llm</span></span></span><br /></dt>
|
||
<dd><dl class="cpp type">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="namespacetensorrt__llm_1_1runtime"></span><span class="k"><span class="pre">namespace</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">runtime</span></span></span><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfigE"></span><span id="tensorrt_llm::runtime::SamplingConfig"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig"></span><span class="k"><span class="pre">class</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">SamplingConfig</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfigE" title="Link to this definition"></a><br /></dt>
|
||
<dd><div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-functions">Public Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig14SamplingConfigE10SizeType32">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig14SamplingConfigE10SizeType32"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig14SamplingConfigE10SizeType32"></span><span id="tensorrt_llm::runtime::SamplingConfig::SamplingConfig__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1a21ca969b785842a734cb5f972e68706d"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">SamplingConfig</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">beamWidth</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">1</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig14SamplingConfigE10SizeType32" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig14SamplingConfigERKNSt6vectorI14SamplingConfigEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig14SamplingConfigERKNSt6vectorI14SamplingConfigEE"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig14SamplingConfigERKNSt6vectorI14SamplingConfigEE"></span><span id="tensorrt_llm::runtime::SamplingConfig::SamplingConfig__std::vector:SamplingConfig:CR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1a56ba4c7f8091d42e62e8651078be6f18"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">SamplingConfig</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig14SamplingConfigERKNSt6vectorI14SamplingConfigEE" title="tensorrt_llm::runtime::SamplingConfig::SamplingConfig"><span class="n"><span class="pre">SamplingConfig</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">configs</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig14SamplingConfigERKNSt6vectorI14SamplingConfigEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig14SamplingConfigERKN8executor14SamplingConfigERKNSt8optionalIN8executor25ExternalDraftTokensConfigEEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig14SamplingConfigERKN8executor14SamplingConfigERKNSt8optionalIN8executor25ExternalDraftTokensConfigEEE"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig14SamplingConfigERKN8executor14SamplingConfigERKNSt8optionalIN8executor25ExternalDraftTokensConfigEEE"></span><span id="tensorrt_llm::runtime::SamplingConfig::SamplingConfig__executor::SamplingConfigCR.std::optional:executor::ExternalDraftTokensConfig:CR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1ae9f28a5af03a0c2a782e7c8dc6abf7c3"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">explicit</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">SamplingConfig</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executorE" title="tensorrt_llm::executor"><span class="n"><span class="pre">executor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executor14SamplingConfigE" title="tensorrt_llm::executor::SamplingConfig"><span class="n"><span class="pre">SamplingConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">samplingConfig</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executorE" title="tensorrt_llm::executor"><span class="n"><span class="pre">executor</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm8executor25ExternalDraftTokensConfigE" title="tensorrt_llm::executor::ExternalDraftTokensConfig"><span class="n"><span class="pre">ExternalDraftTokensConfig</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">externalDraftTokensConfig</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig14SamplingConfigERKN8executor14SamplingConfigERKNSt8optionalIN8executor25ExternalDraftTokensConfigEEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig8validateEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig8validateEv"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig8validateEv"></span><span id="tensorrt_llm::runtime::SamplingConfig::validate"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1a1ee5ddc2543fe7b3dc6389b7d8fc763b"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">validate</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig8validateEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig16useDefaultValuesEbRK6OptVecI1TE1T">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime14SamplingConfig16useDefaultValuesERK6OptVecI1TE1T"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime14SamplingConfig16useDefaultValuesERK6OptVecI1TE1T"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1a03ce780bf1e9b1c48793e6b8dff319bd"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">useDefaultValues</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE" title="tensorrt_llm::runtime::SamplingConfig::OptVec"><span class="n"><span class="pre">OptVec</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig16useDefaultValuesEbRK6OptVecI1TE1T" title="tensorrt_llm::runtime::SamplingConfig::useDefaultValues::T"><span class="n"><span class="pre">T</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">vec</span></span>, <a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig16useDefaultValuesEbRK6OptVecI1TE1T" title="tensorrt_llm::runtime::SamplingConfig::useDefaultValues::T"><span class="n"><span class="pre">T</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">defaultValue</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig16useDefaultValuesEbRK6OptVecI1TE1T" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14SamplingConfigeqERK14SamplingConfig">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14SamplingConfigeqERK14SamplingConfig"></span><span id="_CPPv2NK12tensorrt_llm7runtime14SamplingConfigeqERK14SamplingConfig"></span><span id="tensorrt_llm::runtime::SamplingConfig::eq-operator__SamplingConfigCRC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1a7d909d5f82efa13555105b8373cb1451"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="k"><span class="pre">operator</span></span><span class="o"><span class="pre">==</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfigE" title="tensorrt_llm::runtime::SamplingConfig"><span class="n"><span class="pre">SamplingConfig</span></span></a><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">other</span></span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime14SamplingConfigeqERK14SamplingConfig" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14SamplingConfig17getNumReturnBeamsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14SamplingConfig17getNumReturnBeamsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14SamplingConfig17getNumReturnBeamsEv"></span><span id="tensorrt_llm::runtime::SamplingConfig::getNumReturnBeamsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1a5abd37189f9136af94760c5580f249e7"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getNumReturnBeams</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">const</span></span><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime14SamplingConfig17getNumReturnBeamsEv" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-public-members">Public Members</p>
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig9beamWidthE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig9beamWidthE"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig9beamWidthE"></span><span id="tensorrt_llm::runtime::SamplingConfig::beamWidth__SizeType32"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1a52bb6d11985ba57e1227da7d1ecc0fd0"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">beamWidth</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig9beamWidthE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig18numReturnSequencesE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig18numReturnSequencesE"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig18numReturnSequencesE"></span><span id="tensorrt_llm::runtime::SamplingConfig::numReturnSequences__std::optional:SizeType32:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1ae761961bb83f67c77d94b56125d298a6"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">numReturnSequences</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig18numReturnSequencesE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig11temperatureE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig11temperatureE"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig11temperatureE"></span><span id="tensorrt_llm::runtime::SamplingConfig::temperature__OptVec:FloatType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1aa0234e25caafb8f7e2540e635354f1a7"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE" title="tensorrt_llm::runtime::SamplingConfig::OptVec"><span class="n"><span class="pre">OptVec</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig9FloatTypeE" title="tensorrt_llm::runtime::SamplingConfig::FloatType"><span class="n"><span class="pre">FloatType</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">temperature</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig11temperatureE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig19originalTemperatureE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig19originalTemperatureE"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig19originalTemperatureE"></span><span id="tensorrt_llm::runtime::SamplingConfig::originalTemperature__OptVec:FloatType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1aaa22b7ffef3a4e0e15efb990fe8907b2"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE" title="tensorrt_llm::runtime::SamplingConfig::OptVec"><span class="n"><span class="pre">OptVec</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig9FloatTypeE" title="tensorrt_llm::runtime::SamplingConfig::FloatType"><span class="n"><span class="pre">FloatType</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">originalTemperature</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig19originalTemperatureE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig9minLengthE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig9minLengthE"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig9minLengthE"></span><span id="tensorrt_llm::runtime::SamplingConfig::minLength__OptVec:SizeType32:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1a30d0f89f7035ccd82315aa0bc3fd7182"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE" title="tensorrt_llm::runtime::SamplingConfig::OptVec"><span class="n"><span class="pre">OptVec</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">minLength</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig9minLengthE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig17repetitionPenaltyE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig17repetitionPenaltyE"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig17repetitionPenaltyE"></span><span id="tensorrt_llm::runtime::SamplingConfig::repetitionPenalty__OptVec:FloatType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1ab267d53af1ff989a2cda73069dfd42fa"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE" title="tensorrt_llm::runtime::SamplingConfig::OptVec"><span class="n"><span class="pre">OptVec</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig9FloatTypeE" title="tensorrt_llm::runtime::SamplingConfig::FloatType"><span class="n"><span class="pre">FloatType</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">repetitionPenalty</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig17repetitionPenaltyE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig15presencePenaltyE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig15presencePenaltyE"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig15presencePenaltyE"></span><span id="tensorrt_llm::runtime::SamplingConfig::presencePenalty__OptVec:FloatType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1af40f62527a6a93da70def3daafdc8001"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE" title="tensorrt_llm::runtime::SamplingConfig::OptVec"><span class="n"><span class="pre">OptVec</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig9FloatTypeE" title="tensorrt_llm::runtime::SamplingConfig::FloatType"><span class="n"><span class="pre">FloatType</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">presencePenalty</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig15presencePenaltyE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig16frequencyPenaltyE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig16frequencyPenaltyE"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig16frequencyPenaltyE"></span><span id="tensorrt_llm::runtime::SamplingConfig::frequencyPenalty__OptVec:FloatType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1a1d642826eaf790461c458263f2ad90c5"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE" title="tensorrt_llm::runtime::SamplingConfig::OptVec"><span class="n"><span class="pre">OptVec</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig9FloatTypeE" title="tensorrt_llm::runtime::SamplingConfig::FloatType"><span class="n"><span class="pre">FloatType</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">frequencyPenalty</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig16frequencyPenaltyE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig17noRepeatNgramSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig17noRepeatNgramSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig17noRepeatNgramSizeE"></span><span id="tensorrt_llm::runtime::SamplingConfig::noRepeatNgramSize__OptVec:SizeType32:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1aeedb42ce71155887fc3c9aea45a423ce"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE" title="tensorrt_llm::runtime::SamplingConfig::OptVec"><span class="n"><span class="pre">OptVec</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">noRepeatNgramSize</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig17noRepeatNgramSizeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig14outputLogProbsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig14outputLogProbsE"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig14outputLogProbsE"></span><span id="tensorrt_llm::runtime::SamplingConfig::outputLogProbs__OptVec:b:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1ad55655684229d974d259c5222ad613f7"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE" title="tensorrt_llm::runtime::SamplingConfig::OptVec"><span class="n"><span class="pre">OptVec</span></span></a><span class="p"><span class="pre"><</span></span><span class="kt"><span class="pre">bool</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">outputLogProbs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig14outputLogProbsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig11cumLogProbsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig11cumLogProbsE"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig11cumLogProbsE"></span><span id="tensorrt_llm::runtime::SamplingConfig::cumLogProbs__OptVec:b:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1a19bc5c481edee164410a04bacbbe81ed"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE" title="tensorrt_llm::runtime::SamplingConfig::OptVec"><span class="n"><span class="pre">OptVec</span></span></a><span class="p"><span class="pre"><</span></span><span class="kt"><span class="pre">bool</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">cumLogProbs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig11cumLogProbsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig4topKE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig4topKE"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig4topKE"></span><span id="tensorrt_llm::runtime::SamplingConfig::topK__OptVec:SizeType32:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1a89d73ce271859ae8623309e97ef38a99"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE" title="tensorrt_llm::runtime::SamplingConfig::OptVec"><span class="n"><span class="pre">OptVec</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">topK</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig4topKE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig4topPE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig4topPE"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig4topPE"></span><span id="tensorrt_llm::runtime::SamplingConfig::topP__OptVec:FloatType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1a6d960ae9fb60fa44c616cf4b16a6977d"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE" title="tensorrt_llm::runtime::SamplingConfig::OptVec"><span class="n"><span class="pre">OptVec</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig9FloatTypeE" title="tensorrt_llm::runtime::SamplingConfig::FloatType"><span class="n"><span class="pre">FloatType</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">topP</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig4topPE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig10randomSeedE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig10randomSeedE"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig10randomSeedE"></span><span id="tensorrt_llm::runtime::SamplingConfig::randomSeed__OptVec:uint64_t:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1ace2fdcbe3634f654db68096f7d89a494"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE" title="tensorrt_llm::runtime::SamplingConfig::OptVec"><span class="n"><span class="pre">OptVec</span></span></a><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">uint64_t</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">randomSeed</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig10randomSeedE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig9topPDecayE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig9topPDecayE"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig9topPDecayE"></span><span id="tensorrt_llm::runtime::SamplingConfig::topPDecay__OptVec:FloatType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1ad25e0aec9100bf05d100885677dfaa03"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE" title="tensorrt_llm::runtime::SamplingConfig::OptVec"><span class="n"><span class="pre">OptVec</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig9FloatTypeE" title="tensorrt_llm::runtime::SamplingConfig::FloatType"><span class="n"><span class="pre">FloatType</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">topPDecay</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig9topPDecayE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig7topPMinE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig7topPMinE"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig7topPMinE"></span><span id="tensorrt_llm::runtime::SamplingConfig::topPMin__OptVec:FloatType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1a2f772d5c0be4d9fe80fdf0e142a731d5"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE" title="tensorrt_llm::runtime::SamplingConfig::OptVec"><span class="n"><span class="pre">OptVec</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig9FloatTypeE" title="tensorrt_llm::runtime::SamplingConfig::FloatType"><span class="n"><span class="pre">FloatType</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">topPMin</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig7topPMinE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig12topPResetIdsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig12topPResetIdsE"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig12topPResetIdsE"></span><span id="tensorrt_llm::runtime::SamplingConfig::topPResetIds__OptVec:TokenIdType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1afffb3130864d729191225855b3144d94"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE" title="tensorrt_llm::runtime::SamplingConfig::OptVec"><span class="n"><span class="pre">OptVec</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11TokenIdTypeE" title="tensorrt_llm::runtime::TokenIdType"><span class="n"><span class="pre">TokenIdType</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">topPResetIds</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig12topPResetIdsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig23beamSearchDiversityRateE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig23beamSearchDiversityRateE"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig23beamSearchDiversityRateE"></span><span id="tensorrt_llm::runtime::SamplingConfig::beamSearchDiversityRate__OptVec:FloatType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1a58a44a42e3086649b6b510222b007ac6"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE" title="tensorrt_llm::runtime::SamplingConfig::OptVec"><span class="n"><span class="pre">OptVec</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig9FloatTypeE" title="tensorrt_llm::runtime::SamplingConfig::FloatType"><span class="n"><span class="pre">FloatType</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">beamSearchDiversityRate</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig23beamSearchDiversityRateE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig13lengthPenaltyE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig13lengthPenaltyE"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig13lengthPenaltyE"></span><span id="tensorrt_llm::runtime::SamplingConfig::lengthPenalty__OptVec:FloatType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1a393378cbfb39e5a147b88a8601050947"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE" title="tensorrt_llm::runtime::SamplingConfig::OptVec"><span class="n"><span class="pre">OptVec</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig9FloatTypeE" title="tensorrt_llm::runtime::SamplingConfig::FloatType"><span class="n"><span class="pre">FloatType</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">lengthPenalty</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig13lengthPenaltyE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig13earlyStoppingE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig13earlyStoppingE"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig13earlyStoppingE"></span><span id="tensorrt_llm::runtime::SamplingConfig::earlyStopping__OptVec:SizeType32:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1a947d2499aa3f805431f64206052dfdf3"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE" title="tensorrt_llm::runtime::SamplingConfig::OptVec"><span class="n"><span class="pre">OptVec</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">earlyStopping</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig13earlyStoppingE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig24draftAcceptanceThresholdE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig24draftAcceptanceThresholdE"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig24draftAcceptanceThresholdE"></span><span id="tensorrt_llm::runtime::SamplingConfig::draftAcceptanceThreshold__OptVec:FloatType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1a914d673110a9a51924ec03567f2b6fb5"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE" title="tensorrt_llm::runtime::SamplingConfig::OptVec"><span class="n"><span class="pre">OptVec</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig9FloatTypeE" title="tensorrt_llm::runtime::SamplingConfig::FloatType"><span class="n"><span class="pre">FloatType</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">draftAcceptanceThreshold</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig24draftAcceptanceThresholdE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig15topKMedusaHeadsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig15topKMedusaHeadsE"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig15topKMedusaHeadsE"></span><span id="tensorrt_llm::runtime::SamplingConfig::topKMedusaHeads__OptVec:std::vector:runtime::SizeType32::"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1a0485682ca1b2e632a48c87e4e515c167"></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE" title="tensorrt_llm::runtime::SamplingConfig::OptVec"><span class="n"><span class="pre">OptVec</span></span></a><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="executor.html#_CPPv4N12tensorrt_llm7runtimeE" title="tensorrt_llm::runtime"><span class="n"><span class="pre">runtime</span></span></a><span class="p"><span class="pre">::</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10SizeType32E" title="tensorrt_llm::runtime::SizeType32"><span class="n"><span class="pre">SizeType32</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">topKMedusaHeads</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig15topKMedusaHeadsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig17normalizeLogProbsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig17normalizeLogProbsE"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig17normalizeLogProbsE"></span><span id="tensorrt_llm::runtime::SamplingConfig::normalizeLogProbs__std::optional:b:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1a7657ff197df68d0b7591497d9434983e"></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><span class="kt"><span class="pre">bool</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">normalizeLogProbs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig17normalizeLogProbsE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-types">Private Types</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14SamplingConfig9FloatTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig9FloatTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig9FloatTypeE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1a78311bb676349c17d54fee63f3d54396"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">FloatType</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="kt"><span class="pre">float</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfig9FloatTypeE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1a82ed7012baf2949351e80937329b530a"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">OptVec</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE" title="tensorrt_llm::runtime::SamplingConfig::OptVec::T"><span class="n"><span class="pre">T</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig3VecE">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime14SamplingConfig3VecE"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime14SamplingConfig3VecE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1aee7951e5eef557e59c10259215085828"></span><span class="k"><span class="pre">using</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">Vec</span></span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig3VecE" title="tensorrt_llm::runtime::SamplingConfig::Vec::T"><span class="n"><span class="pre">T</span></span></a><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig3VecE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-functions">Private Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig11validateVecEbNSt6stringERK6OptVecI1TE1TNSt8optionalI1TEE">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime14SamplingConfig11validateVecENSt6stringERK6OptVecI1TE1TNSt8optionalI1TEE"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime14SamplingConfig11validateVecENSt6stringERK6OptVecI1TE1TNSt8optionalI1TEE"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1aa734148feaaa8708c45664ed0e293811"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="kt"><span class="pre">bool</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">validateVec</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">string</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">name</span></span>, <a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE" title="tensorrt_llm::runtime::SamplingConfig::OptVec"><span class="n"><span class="pre">OptVec</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig11validateVecEbNSt6stringERK6OptVecI1TE1TNSt8optionalI1TEE" title="tensorrt_llm::runtime::SamplingConfig::validateVec::T"><span class="n"><span class="pre">T</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">vec</span></span>, <a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig11validateVecEbNSt6stringERK6OptVecI1TE1TNSt8optionalI1TEE" title="tensorrt_llm::runtime::SamplingConfig::validateVec::T"><span class="n"><span class="pre">T</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">min</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig11validateVecEbNSt6stringERK6OptVecI1TE1TNSt8optionalI1TEE" title="tensorrt_llm::runtime::SamplingConfig::validateVec::T"><span class="n"><span class="pre">T</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="n sig-param"><span class="pre">max</span></span><span class="w"> </span><span class="p"><span class="pre">=</span></span><span class="w"> </span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">nullopt</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig11validateVecEbNSt6stringERK6OptVecI1TE1TNSt8optionalI1TEE" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<div class="breathe-sectiondef docutils container">
|
||
<p class="breathe-sectiondef-title rubric" id="breathe-section-title-private-static-functions">Private Static Functions</p>
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig10fuseValuesE6OptVecI1TERKNSt6vectorI14SamplingConfigEENSt8functionIF6OptVecI1TE6size_tEEE1T">
|
||
<span id="_CPPv3I0EN12tensorrt_llm7runtime14SamplingConfig10fuseValuesERKNSt6vectorI14SamplingConfigEENSt8functionIF6OptVecI1TE6size_tEEE1T"></span><span id="_CPPv2I0EN12tensorrt_llm7runtime14SamplingConfig10fuseValuesERKNSt6vectorI14SamplingConfigEENSt8functionIF6OptVecI1TE6size_tEEE1T"></span><span class="k"><span class="pre">template</span></span><span class="p"><span class="pre"><</span></span><span class="k"><span class="pre">typename</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">T</span></span></span><span class="p"><span class="pre">></span></span><br /><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1aaf40d6f66a7a24f1d0328ea68e104bef"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE" title="tensorrt_llm::runtime::SamplingConfig::OptVec"><span class="n"><span class="pre">OptVec</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig10fuseValuesE6OptVecI1TERKNSt6vectorI14SamplingConfigEENSt8functionIF6OptVecI1TE6size_tEEE1T" title="tensorrt_llm::runtime::SamplingConfig::fuseValues::T"><span class="n"><span class="pre">T</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">fuseValues</span></span></span><span class="sig-paren">(</span><span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">vector</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14SamplingConfigE" title="tensorrt_llm::runtime::SamplingConfig"><span class="n"><span class="pre">SamplingConfig</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">configs</span></span>, <span class="n"><span class="pre">std</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">function</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig6OptVecE" title="tensorrt_llm::runtime::SamplingConfig::OptVec"><span class="n"><span class="pre">OptVec</span></span></a><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig10fuseValuesE6OptVecI1TERKNSt6vectorI14SamplingConfigEENSt8functionIF6OptVecI1TE6size_tEEE1T" title="tensorrt_llm::runtime::SamplingConfig::fuseValues::T"><span class="n"><span class="pre">T</span></span></a><span class="p"><span class="pre">></span></span><span class="p"><span class="pre">(</span></span><span class="n"><span class="pre">size_t</span></span><span class="w"> </span><span class="n"><span class="pre">ci</span></span><span class="p"><span class="pre">)</span></span><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="n sig-param"><span class="pre">accessor</span></span>, <a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig10fuseValuesE6OptVecI1TERKNSt6vectorI14SamplingConfigEENSt8functionIF6OptVecI1TE6size_tEEE1T" title="tensorrt_llm::runtime::SamplingConfig::fuseValues::T"><span class="n"><span class="pre">T</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">defaultValue</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime14SamplingConfig10fuseValuesE6OptVecI1TERKNSt6vectorI14SamplingConfigEENSt8functionIF6OptVecI1TE6size_tEEE1T" title="Link to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
</section>
|
||
|
||
|
||
</div>
|
||
</div>
|
||
<footer><div class="rst-footer-buttons" role="navigation" aria-label="Footer">
|
||
<a href="executor.html" class="btn btn-neutral float-left" title="Executor" accesskey="p" rel="prev"><span class="fa fa-arrow-circle-left" aria-hidden="true"></span> Previous</a>
|
||
<a href="../commands/trtllm-build.html" class="btn btn-neutral float-right" title="trtllm-build" accesskey="n" rel="next">Next <span class="fa fa-arrow-circle-right" aria-hidden="true"></span></a>
|
||
</div>
|
||
|
||
<hr/>
|
||
|
||
<div role="contentinfo">
|
||
<jinja2.runtime.BlockReference object at 0x7f96d21b59a0>
|
||
|
||
<div class="footer">
|
||
<p>
|
||
Copyright © 2024 NVIDIA Corporation
|
||
</p>
|
||
<p>
|
||
<a class="Link" href="https://www.nvidia.com/en-us/about-nvidia/privacy-policy/" target="_blank" rel="noopener"
|
||
data-cms-ai="0">Privacy Policy</a> |
|
||
<a class="Link" href="https://www.nvidia.com/en-us/about-nvidia/privacy-center/" target="_blank" rel="noopener"
|
||
data-cms-ai="0">Manage My Privacy</a> |
|
||
<a class="Link" href="https://www.nvidia.com/en-us/preferences/start/" target="_blank" rel="noopener"
|
||
data-cms-ai="0">Do Not Sell or Share My Data</a> |
|
||
<a class="Link" href="https://www.nvidia.com/en-us/about-nvidia/terms-of-service/" target="_blank"
|
||
rel="noopener" data-cms-ai="0">Terms of Service</a> |
|
||
<a class="Link" href="https://www.nvidia.com/en-us/about-nvidia/accessibility/" target="_blank" rel="noopener"
|
||
data-cms-ai="0">Accessibility</a> |
|
||
<a class="Link" href="https://www.nvidia.com/en-us/about-nvidia/company-policies/" target="_blank"
|
||
rel="noopener" data-cms-ai="0">Corporate Policies</a> |
|
||
<a class="Link" href="https://www.nvidia.com/en-us/product-security/" target="_blank" rel="noopener"
|
||
data-cms-ai="0">Product Security</a> |
|
||
<a class="Link" href="https://www.nvidia.com/en-us/contact/" target="_blank" rel="noopener"
|
||
data-cms-ai="0">Contact</a>
|
||
</p>
|
||
</div>
|
||
|
||
|
||
</div>
|
||
|
||
|
||
|
||
</footer>
|
||
</div>
|
||
</div>
|
||
</section>
|
||
</div>
|
||
<script>
|
||
jQuery(function () {
|
||
SphinxRtdTheme.Navigation.enable(true);
|
||
});
|
||
</script>
|
||
|
||
</body>
|
||
</html> |