mirror of
https://github.com/NVIDIA/TensorRT-LLM.git
synced 2026-01-14 06:27:45 +08:00
4625 lines
953 KiB
HTML
4625 lines
953 KiB
HTML
<!DOCTYPE html>
|
||
<html class="writer-html5" lang="en" >
|
||
<head>
|
||
<meta charset="utf-8" /><meta name="generator" content="Docutils 0.18.1: http://docutils.sourceforge.net/" />
|
||
|
||
<meta name="viewport" content="width=device-width, initial-scale=1.0" />
|
||
<title>Runtime — tensorrt_llm documentation</title>
|
||
<link rel="stylesheet" href="../_static/pygments.css" type="text/css" />
|
||
<link rel="stylesheet" href="../_static/css/theme.css" type="text/css" />
|
||
<!--[if lt IE 9]>
|
||
<script src="../_static/js/html5shiv.min.js"></script>
|
||
<![endif]-->
|
||
|
||
<script src="../_static/jquery.js?v=5d32c60e"></script>
|
||
<script src="../_static/_sphinx_javascript_frameworks_compat.js?v=2cd50e6c"></script>
|
||
<script data-url_root="../" id="documentation_options" src="../_static/documentation_options.js?v=b3ba4146"></script>
|
||
<script src="../_static/doctools.js?v=888ff710"></script>
|
||
<script src="../_static/sphinx_highlight.js?v=4825356b"></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="prev" title="Runtime" href="../python-api/tensorrt_llm.runtime.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">Contents:</span></p>
|
||
<ul>
|
||
<li class="toctree-l1"><a class="reference internal" href="../architecture.html">TensorRT-LLM Architecture</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../gpt_runtime.html">C++ GPT Runtime</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../batch_manager.html">The Batch Manager in TensorRT-LLM</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../gpt_attention.html">Multi-head, Multi-query and Group-query Attention</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../precision.html">Numerical Precision</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../performance.html">Performance of TensorRT-LLM</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../installation.html">Build From Sources</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../2023-05-19-how-to-debug.html">How to debug</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../2023-05-17-how-to-add-a-new-model.html">How to add a new model</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../graph-rewriting.html">Graph Rewriting Module</a></li>
|
||
</ul>
|
||
<p class="caption" role="heading"><span class="caption-text">Python 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">Qunatization</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 current"><a class="current reference internal" href="#">Runtime</a><ul>
|
||
<li class="toctree-l2"><a class="reference internal" href="#buffermanager-h">bufferManager.h</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#common-h">common.h</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#cudaevent-h">cudaEvent.h</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#cudastream-h">cudaStream.h</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#decodinginput-h">decodingInput.h</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#decodingoutput-h">decodingOutput.h</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#generationinput-h">generationInput.h</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#generationoutput-h">generationOutput.h</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#gptdecoder-h">gptDecoder.h</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#gptdecoderbatch-h">gptDecoderBatch.h</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#gptjsonconfig-h">gptJsonConfig.h</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#gptmodelconfig-h">gptModelConfig.h</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#gptsession-h">gptSession.h</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#ibuffer-h">iBuffer.h</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#igptdecoderbatch-h">iGptDecoderBatch.h</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#istatefulgptdecoder-h">iStatefulGptDecoder.h</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#itensor-h">iTensor.h</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#ipcutils-h">ipcUtils.h</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#memorycounters-h">memoryCounters.h</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#samplingconfig-h">samplingConfig.h</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#tllmlogger-h">tllmLogger.h</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="#worldconfig-h">worldConfig.h</a></li>
|
||
</ul>
|
||
</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="Permalink to this heading"></a></h1>
|
||
<section id="buffermanager-h">
|
||
<h2>bufferManager.h<a class="headerlink" href="#buffermanager-h" title="Permalink 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><a class="headerlink" href="#_CPPv412tensorrt_llm" title="Permalink to this definition"></a><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><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtimeE" title="Permalink to this definition"></a><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="Permalink 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="Permalink 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="Permalink 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="Permalink 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_llm7runtime13BufferManager13BufferManagerE13CudaStreamPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13BufferManager13BufferManagerE13CudaStreamPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime13BufferManager13BufferManagerE13CudaStreamPtr"></span><span id="tensorrt_llm::runtime::BufferManager::BufferManager__CudaStreamPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1ac641ea2cdcc436ed29ce3bacc42d549a"></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="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager13BufferManagerE13CudaStreamPtr" title="Permalink 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="_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="Permalink 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. </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="Permalink 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. </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="Permalink 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="Permalink 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="Permalink 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="Permalink 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_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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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_1aa8eb6eafae45f681f082cab5daa26029"></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="Permalink to this definition"></a><br /></dt>
|
||
<dd><p>Get the underlying cuda 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_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="Permalink 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="Permalink 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="Permalink 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="Permalink 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>
|
||
|
||
</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_1af4254e3a9525a8051face3c7ca729b65"></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">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="Permalink 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="Permalink 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="_CPPv4N12tensorrt_llm7runtime13BufferManager14initMemoryPoolEi">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13BufferManager14initMemoryPoolEi"></span><span id="_CPPv2N12tensorrt_llm7runtime13BufferManager14initMemoryPoolEi"></span><span id="tensorrt_llm::runtime::BufferManager::initMemoryPool__i"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferManager_1aba8fe4df93bbf829d0cefa8bcc63e657"></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">initMemoryPool</span></span></span><span class="sig-paren">(</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="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13BufferManager14initMemoryPoolEi" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="common-h">
|
||
<h2>common.h<a class="headerlink" href="#common-h" title="Permalink 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-typedefs">Typedefs</p>
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime8SizeTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime8SizeTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime8SizeTypeE"></span><span class="target" id="common_8h_1a8da5a8de61b5f096333bdd15d5a16875"></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">SizeType</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_llm7runtime8SizeTypeE" title="Permalink 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="Permalink 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="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="cudaevent-h">
|
||
<h2>cudaEvent.h<a class="headerlink" href="#cudaevent-h" title="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="_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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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_llm7runtime13DecodingInput9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1a8e97b9538b8259558b0c15cbe2a59d15"></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="w"> </span><span class="k"><span class="pre">const</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput9TensorPtrE" title="Permalink 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_llm7runtime13DecodingInput13DecodingInputE8SizeType8SizeType9TensorPtr9TensorPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput13DecodingInputE8SizeType8SizeType9TensorPtr9TensorPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput13DecodingInputE8SizeType8SizeType9TensorPtr9TensorPtr"></span><span id="tensorrt_llm::runtime::DecodingInput::DecodingInput__SizeType.SizeType.TensorPtr.TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1a4dc28cfb590912072e86070136f4e40a"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxLength</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchSize</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">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><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput13DecodingInputE8SizeType8SizeType9TensorPtr9TensorPtr" title="Permalink 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__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1a82755273657729ca89f9186abeeb713c"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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="Permalink to this definition"></a><br /></dt>
|
||
<dd></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__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1ac4e6e64f43c3994f785a7047afd4578e"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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="Permalink to this definition"></a><br /></dt>
|
||
<dd></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__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1a5d6979ade3dcff577fcf46c70c887917"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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="Permalink to this definition"></a><br /></dt>
|
||
<dd></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__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1aafadc431e8d06615d8425d82b1afaaa8"></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">logits</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput6logitsE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></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__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1a6bb630897ccbb4f437cf9ec517912d25"></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">endIds</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput6endIdsE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></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__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1aa14e54120b58626beca8705ced84a79c"></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">sequenceLimitLength</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput19sequenceLimitLengthE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></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__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1a132486f973befa6c1c54bb0c22aab289"></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">embeddingBias</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput13embeddingBiasE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></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__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1ad44428022172d330f631c81a712ce206"></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">lengths</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput7lengthsE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput12badWordsListE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput12badWordsListE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput12badWordsListE"></span><span id="tensorrt_llm::runtime::DecodingInput::badWordsList__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1acfc2ebcd8202fc2330466a8773b1ab6d"></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">badWordsList</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput12badWordsListE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13DecodingInput13stopWordsListE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13DecodingInput13stopWordsListE"></span><span id="_CPPv2N12tensorrt_llm7runtime13DecodingInput13stopWordsListE"></span><span id="tensorrt_llm::runtime::DecodingInput::stopWordsList__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1a6979dbc3a119fefd9f50c2132c1c95e7"></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">stopWordsList</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput13stopWordsListE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></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__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingInput_1a764f889cfac79873d7c88978e9d42df4"></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">noRepeatNgramSize</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13DecodingInput17noRepeatNgramSizeE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></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="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="decodingoutput-h">
|
||
<h2>decodingOutput.h<a class="headerlink" href="#decodingoutput-h" title="Permalink 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_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="Permalink 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="Permalink 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_llm7runtime14DecodingOutput14DecodingOutputE9TensorPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput14DecodingOutputE9TensorPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput14DecodingOutputE9TensorPtr"></span><span id="tensorrt_llm::runtime::DecodingOutput::DecodingOutput__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1a33bcc26e94af9077c4c8d1ba65acd39c"></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><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14DecodingOutputE9TensorPtr" title="Permalink 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="Permalink 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="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput8finishedE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput8finishedE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput8finishedE"></span><span id="tensorrt_llm::runtime::DecodingOutput::finished__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1a721bfb18d093d881f2f3a7d288dfb0e4"></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">finished</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput8finishedE" title="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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_1afad8e90f48ab5da093fdbb84b801f6da"></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">float</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="Permalink 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="Permalink 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="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses7reshapeE8SizeType8SizeType8SizeType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses7reshapeE8SizeType8SizeType8SizeType"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses7reshapeE8SizeType8SizeType8SizeType"></span><span id="tensorrt_llm::runtime::DecodingOutput::BeamHypotheses::reshape__SizeType.SizeType.SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1BeamHypotheses_1a5ca4a3c3e34fe1f94224866821a2c352"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">beamWidth</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime14DecodingOutput14BeamHypotheses7reshapeE8SizeType8SizeType8SizeType" title="Permalink 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="Permalink 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="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses5sliceE8SizeType8SizeType">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses5sliceE8SizeType8SizeType"></span><span id="_CPPv2NK12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses5sliceE8SizeType8SizeType"></span><span id="tensorrt_llm::runtime::DecodingOutput::BeamHypotheses::slice__SizeType.SizeTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1BeamHypotheses_1a35faed30b9315b8e1890c64d77561ac1"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchIndex</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime14DecodingOutput14BeamHypotheses5sliceE8SizeType8SizeType" title="Permalink 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_llm7runtime14DecodingOutput14BeamHypotheses12outputIdsTgtE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses12outputIdsTgtE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses12outputIdsTgtE"></span><span id="tensorrt_llm::runtime::DecodingOutput::BeamHypotheses::outputIdsTgt__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1BeamHypotheses_1aaeebc4faabecf671bd786c5ff44e2364"></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">outputIdsTgt</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses12outputIdsTgtE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses18sequenceLengthsTgtE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses18sequenceLengthsTgtE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses18sequenceLengthsTgtE"></span><span id="tensorrt_llm::runtime::DecodingOutput::BeamHypotheses::sequenceLengthsTgt__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1BeamHypotheses_1a42ed879ae2b5475b54aa7d3328b5357b"></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">sequenceLengthsTgt</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses18sequenceLengthsTgtE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses11cumLogProbsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses11cumLogProbsE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses11cumLogProbsE"></span><span id="tensorrt_llm::runtime::DecodingOutput::BeamHypotheses::cumLogProbs__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1BeamHypotheses_1ad2aaa38891d139affc6713f4d1a6e207"></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_llm7runtime14DecodingOutput14BeamHypotheses11cumLogProbsE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses12normedScoresE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses12normedScoresE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses12normedScoresE"></span><span id="tensorrt_llm::runtime::DecodingOutput::BeamHypotheses::normedScores__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1BeamHypotheses_1af97b849bf039b95020a3e9d7601b77f0"></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">normedScores</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses12normedScoresE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses8logProbsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses8logProbsE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses8logProbsE"></span><span id="tensorrt_llm::runtime::DecodingOutput::BeamHypotheses::logProbs__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1BeamHypotheses_1a51b9823628bb1388d261bbd18b4e141e"></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_llm7runtime14DecodingOutput14BeamHypotheses8logProbsE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses15minNormedScoresE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses15minNormedScoresE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses15minNormedScoresE"></span><span id="tensorrt_llm::runtime::DecodingOutput::BeamHypotheses::minNormedScores__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1BeamHypotheses_1a545ce50c9d6ba29ebd84bbe1c7379648"></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">minNormedScores</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses15minNormedScoresE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses8numBeamsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses8numBeamsE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses8numBeamsE"></span><span id="tensorrt_llm::runtime::DecodingOutput::BeamHypotheses::numBeams__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1BeamHypotheses_1a207a33747d2474bc9252b6105e1cf7e9"></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">numBeams</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses8numBeamsE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses6isDoneE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses6isDoneE"></span><span id="_CPPv2N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses6isDoneE"></span><span id="tensorrt_llm::runtime::DecodingOutput::BeamHypotheses::isDone__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1DecodingOutput_1_1BeamHypotheses_1a75ad5911c7aa09ed3214d0d2ee2e4bad"></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">isDone</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14DecodingOutput14BeamHypotheses6isDoneE" title="Permalink 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="Permalink 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><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15GenerationInputE" title="Permalink 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_llm7runtime15GenerationInput9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GenerationInput9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GenerationInput9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenerationInput_1ad4d928e8ccca5918cf0f158377a91334"></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_llm7runtime15GenerationInput9TensorPtrE" title="Permalink 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_llm7runtime15GenerationInput15GenerationInputEK8SizeTypeK8SizeType9TensorPtr9TensorPtrb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GenerationInput15GenerationInputEK8SizeTypeK8SizeType9TensorPtr9TensorPtrb"></span><span id="_CPPv2N12tensorrt_llm7runtime15GenerationInput15GenerationInputEK8SizeTypeK8SizeType9TensorPtr9TensorPtrb"></span><span id="tensorrt_llm::runtime::GenerationInput::GenerationInput__SizeTypeC.SizeTypeC.TensorPtr.TensorPtr.b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenerationInput_1a919fd8de0e3905a059b746f1e9e35dbe"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime15GenerationInput15GenerationInputEK8SizeTypeK8SizeType9TensorPtr9TensorPtrb" title="Permalink 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_llm7runtime15GenerationInput5endIdE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GenerationInput5endIdE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GenerationInput5endIdE"></span><span id="tensorrt_llm::runtime::GenerationInput::endId__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenerationInput_1a45567fc133775036062587d9ece61eb8"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime15GenerationInput5endIdE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GenerationInput5padIdE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GenerationInput5padIdE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GenerationInput5padIdE"></span><span id="tensorrt_llm::runtime::GenerationInput::padId__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenerationInput_1adf789d32814e781d64e3abe485acc36a"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime15GenerationInput5padIdE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GenerationInput3idsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GenerationInput3idsE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GenerationInput3idsE"></span><span id="tensorrt_llm::runtime::GenerationInput::ids__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenerationInput_1aa8bd15f4c48f06d5270c4bd860b16c00"></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="sig-name descname"><span class="n"><span class="pre">ids</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15GenerationInput3idsE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GenerationInput7lengthsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GenerationInput7lengthsE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GenerationInput7lengthsE"></span><span id="tensorrt_llm::runtime::GenerationInput::lengths__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenerationInput_1a6e9afbe2aabafc722c81144d258e88fa"></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="sig-name descname"><span class="n"><span class="pre">lengths</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15GenerationInput7lengthsE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GenerationInput6packedE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GenerationInput6packedE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GenerationInput6packedE"></span><span id="tensorrt_llm::runtime::GenerationInput::packed__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenerationInput_1ae16b0e7b06def770e9e0037532d2bfe2"></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_llm7runtime15GenerationInput6packedE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GenerationInput16embeddingBiasOptE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GenerationInput16embeddingBiasOptE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GenerationInput16embeddingBiasOptE"></span><span id="tensorrt_llm::runtime::GenerationInput::embeddingBiasOpt__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenerationInput_1a0976efe9d64d9aeb7a6af8acadf32958"></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="sig-name descname"><span class="n"><span class="pre">embeddingBiasOpt</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15GenerationInput16embeddingBiasOptE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GenerationInput12badWordsListE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GenerationInput12badWordsListE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GenerationInput12badWordsListE"></span><span id="tensorrt_llm::runtime::GenerationInput::badWordsList__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenerationInput_1ae9a6430ce531bfc02be2ee95b1809e5b"></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="sig-name descname"><span class="n"><span class="pre">badWordsList</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15GenerationInput12badWordsListE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GenerationInput13stopWordsListE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GenerationInput13stopWordsListE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GenerationInput13stopWordsListE"></span><span id="tensorrt_llm::runtime::GenerationInput::stopWordsList__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenerationInput_1afa302004663b93df44247bedc31d1fa1"></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="sig-name descname"><span class="n"><span class="pre">stopWordsList</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15GenerationInput13stopWordsListE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GenerationInput12maxNewTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GenerationInput12maxNewTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GenerationInput12maxNewTokensE"></span><span id="tensorrt_llm::runtime::GenerationInput::maxNewTokens__std::optional:SizeType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenerationInput_1aad080660c7e4cf593d37f62632dbf63b"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime15GenerationInput12maxNewTokensE" title="Permalink 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="Permalink 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><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime16GenerationOutputE" title="Permalink 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_llm7runtime16GenerationOutput9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16GenerationOutput9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime16GenerationOutput9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenerationOutput_1ac2bcdfc419f03bf441a1d863a5032aba"></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_llm7runtime16GenerationOutput9TensorPtrE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime16GenerationOutput8CallbackE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16GenerationOutput8CallbackE"></span><span id="_CPPv2N12tensorrt_llm7runtime16GenerationOutput8CallbackE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenerationOutput_1a039ff4327d4017bde5e6ca805bae887e"></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_llm7runtime16GenerationOutput9TensorPtrE" title="tensorrt_llm::runtime::GenerationOutput::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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime16GenerationOutput8CallbackE" title="Permalink 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_llm7runtime16GenerationOutput16GenerationOutputE9TensorPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16GenerationOutput16GenerationOutputE9TensorPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime16GenerationOutput16GenerationOutputE9TensorPtr"></span><span id="tensorrt_llm::runtime::GenerationOutput::GenerationOutput__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenerationOutput_1a36b18b313b4fe3aa0c765dd99b69fc64"></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><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime16GenerationOutput16GenerationOutputE9TensorPtr" title="Permalink 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_llm7runtime16GenerationOutput3idsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16GenerationOutput3idsE"></span><span id="_CPPv2N12tensorrt_llm7runtime16GenerationOutput3idsE"></span><span id="tensorrt_llm::runtime::GenerationOutput::ids__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenerationOutput_1abe34ae6955cb9b92a3737c3ff8007d31"></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="sig-name descname"><span class="n"><span class="pre">ids</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime16GenerationOutput3idsE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime16GenerationOutput8logProbsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16GenerationOutput8logProbsE"></span><span id="_CPPv2N12tensorrt_llm7runtime16GenerationOutput8logProbsE"></span><span id="tensorrt_llm::runtime::GenerationOutput::logProbs__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenerationOutput_1ad05894034e839afc1a5b136aea1ec5b7"></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="sig-name descname"><span class="n"><span class="pre">logProbs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime16GenerationOutput8logProbsE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime16GenerationOutput13contextLogitsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16GenerationOutput13contextLogitsE"></span><span id="_CPPv2N12tensorrt_llm7runtime16GenerationOutput13contextLogitsE"></span><span id="tensorrt_llm::runtime::GenerationOutput::contextLogits__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenerationOutput_1a2f45fb2c7cf0ab1c4d93098afdf60fa3"></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="sig-name descname"><span class="n"><span class="pre">contextLogits</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime16GenerationOutput13contextLogitsE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime16GenerationOutput16onTokenGeneratedE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16GenerationOutput16onTokenGeneratedE"></span><span id="_CPPv2N12tensorrt_llm7runtime16GenerationOutput16onTokenGeneratedE"></span><span id="tensorrt_llm::runtime::GenerationOutput::onTokenGenerated__Callback"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GenerationOutput_1a01e4b5ca66e302f380c156896849ee91"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16GenerationOutput8CallbackE" title="tensorrt_llm::runtime::GenerationOutput::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_llm7runtime16GenerationOutput16onTokenGeneratedE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="gptdecoder-h">
|
||
<h2>gptDecoder.h<a class="headerlink" href="#gptdecoder-h" title="Permalink 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="Permalink to this definition"></a><br /></dt>
|
||
<dd><dl class="cpp class">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I0EN12tensorrt_llm6layers18DynamicDecodeLayerE">
|
||
<span id="_CPPv3I0EN12tensorrt_llm6layers18DynamicDecodeLayerE"></span><span id="_CPPv2I0EN12tensorrt_llm6layers18DynamicDecodeLayerE"></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_1layers_1_1DynamicDecodeLayer"></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">DynamicDecodeLayer</span></span></span><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm6layers18DynamicDecodeLayerE" title="Permalink to this definition"></a><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="_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="#_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_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="Permalink 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="Permalink 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_llm7runtime10GptDecoder10GptDecoderE6size_t6size_tRK13CudaStreamPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptDecoder10GptDecoderE6size_t6size_tRK13CudaStreamPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptDecoder10GptDecoderE6size_t6size_tRK13CudaStreamPtr"></span><span id="tensorrt_llm::runtime::GptDecoder::GptDecoder__s.s.CudaStreamPtrCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoder_1a0cbdae7bd365d6175f9fe78a0237cc7c"></span><span class="sig-name descname"><span class="n"><span class="pre">GptDecoder</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">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>, <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="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder10GptDecoderE6size_t6size_tRK13CudaStreamPtr" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptDecoder5setupERK14SamplingConfig6size_t">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptDecoder5setupERK14SamplingConfig6size_t"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptDecoder5setupERK14SamplingConfig6size_t"></span><span id="tensorrt_llm::runtime::GptDecoder::setup__SamplingConfigCR.s"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoder_1af426907b7aac4424847ee592c0596585"></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="sig-paren">)</span><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder5setupERK14SamplingConfig6size_t" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptDecoder7forwardER14DecodingOutputRK13DecodingInput">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptDecoder7forwardER14DecodingOutputRK13DecodingInput"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptDecoder7forwardER14DecodingOutputRK13DecodingInput"></span><span id="tensorrt_llm::runtime::GptDecoder::forward__DecodingOutputR.DecodingInputCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoder_1a687dcb28ec2d720be9c35694250d346f"></span><span class="k"><span class="pre">virtual</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">forward</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_llm7runtime10GptDecoder7forwardER14DecodingOutputRK13DecodingInput" title="Permalink 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="Permalink 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__BufferManager"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoder_1a8602244879ca7e9a83ccb7811cea26a8"></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">mManager</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder8mManagerE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptDecoder10mAllocatorE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptDecoder10mAllocatorE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptDecoder10mAllocatorE"></span><span id="tensorrt_llm::runtime::GptDecoder::mAllocator__common::CudaAllocator"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoder_1a51093c91db9709a5cdb1d6004c19ae57"></span><span class="n"><span class="pre">common</span></span><span class="p"><span class="pre">::</span></span><span class="n"><span class="pre">CudaAllocator</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mAllocator</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptDecoder10mAllocatorE" title="Permalink 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="#_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><a class="reference internal" href="#_CPPv4I0EN12tensorrt_llm6layers18DynamicDecodeLayerE" title="tensorrt_llm::layers::DynamicDecodeLayer"><span class="n"><span class="pre">DynamicDecodeLayer</span></span></a><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="Permalink 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="Permalink 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-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="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11IGptDecoder5setupERK14SamplingConfig6size_t">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11IGptDecoder5setupERK14SamplingConfig6size_t"></span><span id="_CPPv2N12tensorrt_llm7runtime11IGptDecoder5setupERK14SamplingConfig6size_t"></span><span id="tensorrt_llm::runtime::IGptDecoder::setup__SamplingConfigCR.s"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoder_1a3dffd84af10329dbc69291086ac931e4"></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="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_t" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11IGptDecoder7forwardER14DecodingOutputRK13DecodingInput">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11IGptDecoder7forwardER14DecodingOutputRK13DecodingInput"></span><span id="_CPPv2N12tensorrt_llm7runtime11IGptDecoder7forwardER14DecodingOutputRK13DecodingInput"></span><span id="tensorrt_llm::runtime::IGptDecoder::forward__DecodingOutputR.DecodingInputCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoder_1a892c655936e25a38a7f0737fbfadfc1d"></span><span class="k"><span class="pre">virtual</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">forward</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_llm7runtime11IGptDecoder7forwardER14DecodingOutputRK13DecodingInput" title="Permalink 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="Permalink 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_llm7runtime11IGptDecoder10gatherTreeER7ITensorRK14DecodingOutputRK13DecodingInputRK13BufferManager">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11IGptDecoder10gatherTreeER7ITensorRK14DecodingOutputRK13DecodingInputRK13BufferManager"></span><span id="_CPPv2N12tensorrt_llm7runtime11IGptDecoder10gatherTreeER7ITensorRK14DecodingOutputRK13DecodingInputRK13BufferManager"></span><span id="tensorrt_llm::runtime::IGptDecoder::gatherTree__ITensorR.DecodingOutputCR.DecodingInputCR.BufferManagerCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoder_1ae754c28ed1ed7f31726ccbc6ebe385ab"></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">gatherTree</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">finalOutputIds</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="w"> </span><span class="k"><span class="pre">const</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">decodingOutput</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">decodingInput</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><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11IGptDecoder10gatherTreeER7ITensorRK14DecodingOutputRK13DecodingInputRK13BufferManager" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11IGptDecoder6createEN8nvinfer18DataTypeE6size_t6size_tRKN13BufferManager13CudaStreamPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11IGptDecoder6createEN8nvinfer18DataTypeE6size_t6size_tRKN13BufferManager13CudaStreamPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime11IGptDecoder6createEN8nvinfer18DataTypeE6size_t6size_tRKN13BufferManager13CudaStreamPtrE"></span><span id="tensorrt_llm::runtime::IGptDecoder::create__nvinfer1::DataType.s.s.BufferManager::CudaStreamPtrCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoder_1af9c668b9e5cdd2e0a65bfd55ba9b8b98"></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="#_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">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>, <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="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11IGptDecoder6createEN8nvinfer18DataTypeE6size_t6size_tRKN13BufferManager13CudaStreamPtrE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="gptdecoderbatch-h">
|
||
<h2>gptDecoderBatch.h<a class="headerlink" href="#gptdecoderbatch-h" title="Permalink 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_llm7runtime15GptDecoderBatchE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatchE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatchE"></span><span id="tensorrt_llm::runtime::GptDecoderBatch"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch"></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">GptDecoderBatch</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="#_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_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_llm7runtime16IGptDecoderBatchE" title="tensorrt_llm::runtime::IGptDecoderBatch"><span class="n"><span class="pre">IGptDecoderBatch</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15GptDecoderBatchE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd><div class="docutils container">
|
||
<em>#include <gptDecoderBatch.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 type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch13CudaStreamPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch13CudaStreamPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch13CudaStreamPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a8206454db9d639c251910c03681a9ed0"></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_llm7runtime15GptDecoderBatch13CudaStreamPtrE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1afda009f36f02088447dc1da45e5e94f4"></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_llm7runtime15GptDecoderBatch9TensorPtrE" title="Permalink 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_llm7runtime15GptDecoderBatch15GptDecoderBatchENSt6size_tENSt6size_tE13CudaStreamPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch15GptDecoderBatchENSt6size_tENSt6size_tE13CudaStreamPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch15GptDecoderBatchENSt6size_tENSt6size_tE13CudaStreamPtr"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::GptDecoderBatch__std::s.std::s.CudaStreamPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1af478d2ea92161498aefa5ddb70d36262"></span><span class="sig-name descname"><span class="n"><span class="pre">GptDecoderBatch</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_llm7runtime15GptDecoderBatch13CudaStreamPtrE" title="tensorrt_llm::runtime::GptDecoderBatch::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="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch15GptDecoderBatchENSt6size_tENSt6size_tE13CudaStreamPtr" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch5setupE8SizeType8SizeType8SizeTypeN8nvinfer18DataTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch5setupE8SizeType8SizeType8SizeTypeN8nvinfer18DataTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch5setupE8SizeType8SizeType8SizeTypeN8nvinfer18DataTypeE"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::setup__SizeType.SizeType.SizeType.nvinfer1::DataType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1ac5f1b0f5fb9ee4322b50ace3e7d22301"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxBatchSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxBeamWidth</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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">dtype</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_llm7runtime15GptDecoderBatch5setupE8SizeType8SizeType8SizeTypeN8nvinfer18DataTypeE" title="Permalink 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_1IGptDecoderBatch_1ab7e07276f33bfa6ef3c90baa49ee6eec"><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_llm7runtime15GptDecoderBatch10newRequestE8SizeTypeRKN13decoder_batch7RequestERK14SamplingConfig">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch10newRequestE8SizeTypeRKN13decoder_batch7RequestERK14SamplingConfig"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch10newRequestE8SizeTypeRKN13decoder_batch7RequestERK14SamplingConfig"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::newRequest__SizeType.decoder_batch::RequestCR.SamplingConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a7214c20bc16c253fd2cb199884cd9859"></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">newRequest</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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><span class="w"> </span><span class="k"><span class="pre">override</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch10newRequestE8SizeTypeRKN13decoder_batch7RequestERK14SamplingConfig" title="Permalink to this definition"></a><br /></dt>
|
||
<dd><p>Initialize the decoder at <code class="docutils literal notranslate"><span class="pre">batchIdx</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_llm7runtime15GptDecoderBatch8newBatchERK15GenerationInputRK14SamplingConfig">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch8newBatchERK15GenerationInputRK14SamplingConfig"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch8newBatchERK15GenerationInputRK14SamplingConfig"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::newBatch__GenerationInputCR.SamplingConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a4362ed782250a10e670f1db2b661cd0b"></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_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">override</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch8newBatchERK15GenerationInputRK14SamplingConfig" title="Permalink 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_llm7runtime15GptDecoderBatch12forwardAsyncERN13decoder_batch6OutputERKN13decoder_batch5InputE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch12forwardAsyncERN13decoder_batch6OutputERKN13decoder_batch5InputE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch12forwardAsyncERN13decoder_batch6OutputERKN13decoder_batch5InputE"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::forwardAsync__decoder_batch::OutputR.decoder_batch::InputCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1aa468001b519ad9767958cb8e660faea5"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><span class="n"><span class="pre">TokenPtr</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_llm7runtime15GptDecoderBatch12forwardAsyncERN13decoder_batch6OutputERKN13decoder_batch5InputE" title="Permalink 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_llm7runtime15GptDecoderBatch11forwardSyncERKN13decoder_batch5TokenE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch11forwardSyncERKN13decoder_batch5TokenE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch11forwardSyncERKN13decoder_batch5TokenE"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::forwardSync__decoder_batch::TokenCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a495a39bf71a7d39cd81495834716760f"></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_batch5TokenE" title="tensorrt_llm::runtime::decoder_batch::Token"><span class="n"><span class="pre">Token</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">e</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_llm7runtime15GptDecoderBatch11forwardSyncERKN13decoder_batch5TokenE" title="Permalink 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_llm7runtime15GptDecoderBatch12forwardAsyncERN7decoder6OutputERKN7decoder5InputE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch12forwardAsyncERN7decoder6OutputERKN7decoder5InputE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch12forwardAsyncERN7decoder6OutputERKN7decoder5InputE"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::forwardAsync__decoder::OutputR.decoder::InputCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1ab0220317a75f65030f59878e26e3d737"></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_llm7runtime15GptDecoderBatch12forwardAsyncERN7decoder6OutputERKN7decoder5InputE" title="Permalink 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_llm7runtime15GptDecoderBatch14isFinishedSyncEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch14isFinishedSyncEv"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch14isFinishedSyncEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::isFinishedSync"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1ac9935f86503e3857f8d2563532531e09"></span><span class="k"><span class="pre">virtual</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">isFinishedSync</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_llm7runtime15GptDecoderBatch14isFinishedSyncEv" title="Permalink 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 and return whether all sequences have finished. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime15GptDecoderBatch11getFinishedEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime15GptDecoderBatch11getFinishedEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime15GptDecoderBatch11getFinishedEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::getFinishedC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a4c69dab125603dd2a4fbd9ab043afa30"></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_llm7runtime15GptDecoderBatch11getFinishedEv" title="Permalink 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_llm7runtime15GptDecoderBatch12getOutputIdsE8SizeType">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime15GptDecoderBatch12getOutputIdsE8SizeType"></span><span id="_CPPv2NK12tensorrt_llm7runtime15GptDecoderBatch12getOutputIdsE8SizeType"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::getOutputIds__SizeTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a313a9ae3a464f84398b9ba038b4af554"></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_llm7runtime15GptDecoderBatch9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatch::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">getOutputIds</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime15GptDecoderBatch12getOutputIdsE8SizeType" title="Permalink 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, 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_llm7runtime15GptDecoderBatch12getOutputIdsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime15GptDecoderBatch12getOutputIdsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime15GptDecoderBatch12getOutputIdsEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::getOutputIdsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a445fc8f3f0db689b2166cb15fdc63df2"></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_llm7runtime15GptDecoderBatch9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatch::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">getOutputIds</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_llm7runtime15GptDecoderBatch12getOutputIdsEv" title="Permalink 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 </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime15GptDecoderBatch17getFinalOutputIdsE8SizeType">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime15GptDecoderBatch17getFinalOutputIdsE8SizeType"></span><span id="_CPPv2NK12tensorrt_llm7runtime15GptDecoderBatch17getFinalOutputIdsE8SizeType"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::getFinalOutputIds__SizeTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1ab3bb06c1c21e53f4597a4a4e40605539"></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">tuple</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><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatch::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">getFinalOutputIds</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime15GptDecoderBatch17getFinalOutputIdsE8SizeType" title="Permalink to this definition"></a><br /></dt>
|
||
<dd><p>Execute postProcessRequest and returns OutputIds for request <code class="docutils literal notranslate"><span class="pre">batchIdx</span></code>. Result will only be available after event returned </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><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_llm7runtime15GptDecoderBatch17getFinalOutputIdsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime15GptDecoderBatch17getFinalOutputIdsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime15GptDecoderBatch17getFinalOutputIdsEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::getFinalOutputIdsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a222f485382bc2c429686df944c3556f0"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatch::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">getFinalOutputIds</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_llm7runtime15GptDecoderBatch17getFinalOutputIdsEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd><p>Execute postProcessRequest and returns OutputIds. </p>
|
||
<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 </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime15GptDecoderBatch12getParentIdsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime15GptDecoderBatch12getParentIdsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime15GptDecoderBatch12getParentIdsEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::getParentIdsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a93178ec5242faf2de970d4d017352dae"></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_llm7runtime15GptDecoderBatch9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatch::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_llm7runtime15GptDecoderBatch12getParentIdsEv" title="Permalink 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_llm7runtime15GptDecoderBatch16getFinishedBeamsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime15GptDecoderBatch16getFinishedBeamsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime15GptDecoderBatch16getFinishedBeamsEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::getFinishedBeamsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a9fd74a125803979c32a246d8bab2d188"></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_llm7runtime15GptDecoderBatch9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatch::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">getFinishedBeams</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_llm7runtime15GptDecoderBatch16getFinishedBeamsEv" title="Permalink 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], marks finished requests (per beam), on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime15GptDecoderBatch16getOutputLengthsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime15GptDecoderBatch16getOutputLengthsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime15GptDecoderBatch16getOutputLengthsEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::getOutputLengthsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1ab1b79b788d470a0a295ac4e940ac2338"></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_llm7runtime15GptDecoderBatch9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatch::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">getOutputLengths</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_llm7runtime15GptDecoderBatch16getOutputLengthsEv" title="Permalink 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], total sequence lengths (per beam), on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime15GptDecoderBatch14getCumLogProbsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime15GptDecoderBatch14getCumLogProbsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime15GptDecoderBatch14getCumLogProbsEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::getCumLogProbsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a3c07c5204d41a4dfdf688a4482d510bb"></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_llm7runtime15GptDecoderBatch9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatch::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_llm7runtime15GptDecoderBatch14getCumLogProbsEv" title="Permalink 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_llm7runtime15GptDecoderBatch12getNewTokensEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime15GptDecoderBatch12getNewTokensEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime15GptDecoderBatch12getNewTokensEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::getNewTokensC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a0f899f3a1a1101cf24375fbd68eff1a9"></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_llm7runtime15GptDecoderBatch9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatch::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><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_llm7runtime15GptDecoderBatch12getNewTokensEv" title="Permalink 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], 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_llm7runtime15GptDecoderBatch10getNbStepsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime15GptDecoderBatch10getNbStepsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime15GptDecoderBatch10getNbStepsEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::getNbStepsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a9c9a6a07562637a290ed93e987bd06a7"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime15GptDecoderBatch10getNbStepsEv" title="Permalink 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_llm7runtime15GptDecoderBatch13getNbFinishedEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime15GptDecoderBatch13getNbFinishedEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime15GptDecoderBatch13getNbFinishedEv"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::getNbFinishedC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a7c4d967f7646e5a78f50d4ac6844be93"></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_llm7runtime15GptDecoderBatch9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatch::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_llm7runtime15GptDecoderBatch13getNbFinishedEv" title="Permalink 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>
|
||
|
||
</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_llm7runtime15GptDecoderBatch13GptDecoderPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch13GptDecoderPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch13GptDecoderPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1adbc3e2ab6fcff2a86fee1974fd2ea358"></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_llm7runtime15GptDecoderBatch13GptDecoderPtrE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch16DecodingInputPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch16DecodingInputPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch16DecodingInputPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1aebed7ff7ea091a472908a7d2cc468684"></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_llm7runtime15GptDecoderBatch16DecodingInputPtrE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch17DecodingOutputPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch17DecodingOutputPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch17DecodingOutputPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a8b297105bd88a4066694adf8a91d4073"></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_llm7runtime15GptDecoderBatch17DecodingOutputPtrE" title="Permalink 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_llm7runtime15GptDecoderBatch18postProcessRequestE8SizeType">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime15GptDecoderBatch18postProcessRequestE8SizeType"></span><span id="_CPPv2NK12tensorrt_llm7runtime15GptDecoderBatch18postProcessRequestE8SizeType"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::postProcessRequest__SizeTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a7c4def0f8b1239b53724fbf6cf576b12"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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><a class="headerlink" href="#_CPPv4NK12tensorrt_llm7runtime15GptDecoderBatch18postProcessRequestE8SizeType" title="Permalink to this definition"></a><br /></dt>
|
||
<dd><p>Gather final results for request <code class="docutils literal notranslate"><span class="pre">batchIdx</span></code></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_llm7runtime15GptDecoderBatch10mVocabSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch10mVocabSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch10mVocabSizeE"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::mVocabSize__std::sC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a91f2563dd7fecc08ba0da2fa073d7763"></span><span class="k"><span class="pre">const</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">mVocabSize</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch10mVocabSizeE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch16mVocabSizePaddedE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch16mVocabSizePaddedE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch16mVocabSizePaddedE"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::mVocabSizePadded__std::sC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1af9abcb9a1802cfe924e5b7ac0e03b98c"></span><span class="k"><span class="pre">const</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">mVocabSizePadded</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch16mVocabSizePaddedE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch7mStreamE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch7mStreamE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch7mStreamE"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::mStream__CudaStreamPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a33675102b6406b35993f06c59508b4fd"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch13CudaStreamPtrE" title="tensorrt_llm::runtime::GptDecoderBatch::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_llm7runtime15GptDecoderBatch7mStreamE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch14mBufferManagerE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch14mBufferManagerE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch14mBufferManagerE"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::mBufferManager__BufferManager"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a73a99bf969ac6a13872b1b93f443bc2d"></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_llm7runtime15GptDecoderBatch14mBufferManagerE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch13mForwardTokenE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch13mForwardTokenE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch13mForwardTokenE"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::mForwardToken__TokenPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1abe39442ac571c5827d98ec295bd0513c"></span><span class="n"><span class="pre">TokenPtr</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mForwardToken</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch13mForwardTokenE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch13mForwardEventE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch13mForwardEventE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch13mForwardEventE"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::mForwardEvent__CudaEvent"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a9d42242aa2aac2bffd0387a6272f57f9"></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_llm7runtime15GptDecoderBatch13mForwardEventE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch8mStreamsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch8mStreamsE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch8mStreamsE"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::mStreams__std::vector:CudaStreamPtr:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a5b1a9b00ef30809bef69336e5adfcc6c"></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_llm7runtime15GptDecoderBatch13CudaStreamPtrE" title="tensorrt_llm::runtime::GptDecoderBatch::CudaStreamPtr"><span class="n"><span class="pre">CudaStreamPtr</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">mStreams</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch8mStreamsE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch9mDecodersE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch9mDecodersE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch9mDecodersE"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::mDecoders__std::vector:GptDecoderPtr:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a656cea9abf1ea0c8b3697a79ab51dde9"></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_llm7runtime15GptDecoderBatch13GptDecoderPtrE" title="tensorrt_llm::runtime::GptDecoderBatch::GptDecoderPtr"><span class="n"><span class="pre">GptDecoderPtr</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">mDecoders</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch9mDecodersE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch15mDecodingInputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch15mDecodingInputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch15mDecodingInputsE"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::mDecodingInputs__std::vector:DecodingInputPtr:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1aeb33339a3b75a2152134fdacd2660786"></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_llm7runtime15GptDecoderBatch16DecodingInputPtrE" title="tensorrt_llm::runtime::GptDecoderBatch::DecodingInputPtr"><span class="n"><span class="pre">DecodingInputPtr</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">mDecodingInputs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch15mDecodingInputsE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch16mDecodingOutputsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch16mDecodingOutputsE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch16mDecodingOutputsE"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::mDecodingOutputs__std::vector:DecodingOutputPtr:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1ab1c5ab4afe168288cb33331a486d5be8"></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_llm7runtime15GptDecoderBatch17DecodingOutputPtrE" title="tensorrt_llm::runtime::GptDecoderBatch::DecodingOutputPtr"><span class="n"><span class="pre">DecodingOutputPtr</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">mDecodingOutputs</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch16mDecodingOutputsE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch19mJointDecodingInputE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch19mJointDecodingInputE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch19mJointDecodingInputE"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::mJointDecodingInput__DecodingInputPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a73642bc20fd9bcd68b262edd55d4d2a7"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch16DecodingInputPtrE" title="tensorrt_llm::runtime::GptDecoderBatch::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_llm7runtime15GptDecoderBatch19mJointDecodingInputE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch20mJointDecodingOutputE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch20mJointDecodingOutputE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch20mJointDecodingOutputE"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::mJointDecodingOutput__DecodingOutputPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a09890f57577749180eb23ba792989e96"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch17DecodingOutputPtrE" title="tensorrt_llm::runtime::GptDecoderBatch::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_llm7runtime15GptDecoderBatch20mJointDecodingOutputE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch8mNbStepsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch8mNbStepsE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch8mNbStepsE"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::mNbSteps__std::vector:SizeType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1ab82d738a6ee4eb24160c4203f50ba47b"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime15GptDecoderBatch8mNbStepsE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch9mFinishedE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch9mFinishedE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch9mFinishedE"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::mFinished__std::vector:b:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a4bacfb74153463dd2ea29e9e58d62139"></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_llm7runtime15GptDecoderBatch9mFinishedE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch12mFinishedSumE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch12mFinishedSumE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch12mFinishedSumE"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::mFinishedSum__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a5e29dd7a16e6dcfdbc6af1f5a30d454d"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch9TensorPtrE" title="tensorrt_llm::runtime::GptDecoderBatch::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_llm7runtime15GptDecoderBatch12mFinishedSumE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch13mMaxNewTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch13mMaxNewTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch13mMaxNewTokensE"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::mMaxNewTokens__std::vector:SizeType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1adda26d8626cb810bd54f0eae099f81d6"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime15GptDecoderBatch13mMaxNewTokensE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch11mBeamWidthsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch11mBeamWidthsE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch11mBeamWidthsE"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::mBeamWidths__std::vector:SizeType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a62af2790978fdbdce912bb55d1538c7a"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime15GptDecoderBatch11mBeamWidthsE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch18mMaxSequenceLengthE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch18mMaxSequenceLengthE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch18mMaxSequenceLengthE"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::mMaxSequenceLength__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a17b9d45dd6c74bddca33b2edb555fe3c"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime15GptDecoderBatch18mMaxSequenceLengthE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime15GptDecoderBatch16mActualBatchSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime15GptDecoderBatch16mActualBatchSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime15GptDecoderBatch16mActualBatchSizeE"></span><span id="tensorrt_llm::runtime::GptDecoderBatch::mActualBatchSize__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptDecoderBatch_1a7e2f18c7b2ebe3d76b568c46380e92c4"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime15GptDecoderBatch16mActualBatchSizeE" title="Permalink 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="Permalink 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="Permalink 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_llm7runtime13GptJsonConfig13GptJsonConfigENSt6stringENSt6stringE8SizeType8SizeTypeRK14GptModelConfig">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13GptJsonConfig13GptJsonConfigENSt6stringENSt6stringE8SizeType8SizeTypeRK14GptModelConfig"></span><span id="_CPPv2N12tensorrt_llm7runtime13GptJsonConfig13GptJsonConfigENSt6stringENSt6stringE8SizeType8SizeTypeRK14GptModelConfig"></span><span id="tensorrt_llm::runtime::GptJsonConfig::GptJsonConfig__ss.ss.SizeType.SizeType.GptModelConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1a66216254b0b7a635b772546973578fd8"></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">precision</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tensorParallelism</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">pipelineParallelism</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14GptModelConfigE" title="tensorrt_llm::runtime::GptModelConfig"><span class="n"><span class="pre">GptModelConfig</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_llm7runtime13GptJsonConfig13GptJsonConfigENSt6stringENSt6stringE8SizeType8SizeTypeRK14GptModelConfig" title="Permalink 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_1a124fb0206f4ac3ff47629bcb8147589d"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14GptModelConfigE" title="tensorrt_llm::runtime::GptModelConfig"><span class="n"><span class="pre">GptModelConfig</span></span></a><span class="w"> </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="Permalink 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_1a929bf268a1dd42fa06b6d05834bd6646"></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="Permalink 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_1ab5c3f3e5854ebdbd1b0f0d26fa8c2a6f"></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="Permalink 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_1af3d5d313552e8497406ce3ae6b18bc11"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><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="Permalink 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_1a2ec48773fe7b568d771f53770d9b6b0d"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><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="Permalink 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_1a589461cb3bd53963212f1eda119d7f2e"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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_1a63163fef3a6b49203c7cf6cbb0855971"></span><span class="k"><span class="pre">const</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">mName</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig5mNameE" title="Permalink 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_1ad17448c5f6c59a7832c433ceec3687ff"></span><span class="k"><span class="pre">const</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">mPrecision</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig10mPrecisionE" title="Permalink 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__SizeTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1a8f0596343a2ad81565cea45b5ac7f6fc"></span><span class="k"><span class="pre">const</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime13GptJsonConfig18mTensorParallelismE" title="Permalink 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__SizeTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1a69e73e4c48faf1f76ebfc16bcb544431"></span><span class="k"><span class="pre">const</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime13GptJsonConfig20mPipelineParallelismE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13GptJsonConfig15mGptModelConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13GptJsonConfig15mGptModelConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime13GptJsonConfig15mGptModelConfigE"></span><span id="tensorrt_llm::runtime::GptJsonConfig::mGptModelConfig__GptModelConfigC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptJsonConfig_1a573b497d8cd2a5197db8544a102c7af0"></span><span class="k"><span class="pre">const</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14GptModelConfigE" title="tensorrt_llm::runtime::GptModelConfig"><span class="n"><span class="pre">GptModelConfig</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mGptModelConfig</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13GptJsonConfig15mGptModelConfigE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="gptmodelconfig-h">
|
||
<h2>gptModelConfig.h<a class="headerlink" href="#gptmodelconfig-h" title="Permalink 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_llm7runtime14GptModelConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfigE"></span><span id="tensorrt_llm::runtime::GptModelConfig"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig"></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">GptModelConfig</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14GptModelConfigE" title="Permalink 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_llm7runtime14GptModelConfig12ModelVariantE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig12ModelVariantE"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig12ModelVariantE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a5ab7e0ffcb31d8d6d19afd863370a635"></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_llm7runtime14GptModelConfig12ModelVariantE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd><p><em>Values:</em></p>
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig12ModelVariant4kGptE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig12ModelVariant4kGptE"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig12ModelVariant4kGptE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a5ab7e0ffcb31d8d6d19afd863370a635aeb854c1219d60212d16ad06f0d454dc8"></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_llm7runtime14GptModelConfig12ModelVariant4kGptE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp enumerator">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig12ModelVariant4kGlmE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig12ModelVariant4kGlmE"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig12ModelVariant4kGlmE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a5ab7e0ffcb31d8d6d19afd863370a635a30a41043cc5346b26f02b1a3e1cf54d5"></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_llm7runtime14GptModelConfig12ModelVariant4kGlmE" title="Permalink 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_llm7runtime14GptModelConfig14GptModelConfigE8SizeType8SizeType8SizeType8SizeTypeN8nvinfer18DataTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig14GptModelConfigE8SizeType8SizeType8SizeType8SizeTypeN8nvinfer18DataTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig14GptModelConfigE8SizeType8SizeType8SizeType8SizeTypeN8nvinfer18DataTypeE"></span><span id="tensorrt_llm::runtime::GptModelConfig::GptModelConfig__SizeType.SizeType.SizeType.SizeType.nvinfer1::DataTypeCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a93a4b8bc5df2e1a25d4b876fdefb1ff6"></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">GptModelConfig</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">vocabSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">nbLayers</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">nbHeads</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime14GptModelConfig14GptModelConfigE8SizeType8SizeType8SizeType8SizeTypeN8nvinfer18DataTypeE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14GptModelConfig12getVocabSizeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14GptModelConfig12getVocabSizeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14GptModelConfig12getVocabSizeEv"></span><span id="tensorrt_llm::runtime::GptModelConfig::getVocabSizeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1aff992ef62ed1999326b48800913b2bee"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><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_llm7runtime14GptModelConfig12getVocabSizeEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14GptModelConfig18getVocabSizePaddedE8SizeType">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14GptModelConfig18getVocabSizePaddedE8SizeType"></span><span id="_CPPv2NK12tensorrt_llm7runtime14GptModelConfig18getVocabSizePaddedE8SizeType"></span><span id="tensorrt_llm::runtime::GptModelConfig::getVocabSizePadded__SizeTypeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1acf1d9e9818c254ee026c012388025480"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime14GptModelConfig18getVocabSizePaddedE8SizeType" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14GptModelConfig11getNbLayersE8SizeType">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14GptModelConfig11getNbLayersE8SizeType"></span><span id="_CPPv2NK12tensorrt_llm7runtime14GptModelConfig11getNbLayersE8SizeType"></span><span id="tensorrt_llm::runtime::GptModelConfig::getNbLayers__SizeTypeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a527a93c79413842c6706295232b0b9ed"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime14GptModelConfig11getNbLayersE8SizeType" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14GptModelConfig10getNbHeadsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14GptModelConfig10getNbHeadsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14GptModelConfig10getNbHeadsEv"></span><span id="tensorrt_llm::runtime::GptModelConfig::getNbHeadsCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a00ef36e3eb9bd188a911b7f80c345e0f"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><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_llm7runtime14GptModelConfig10getNbHeadsEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14GptModelConfig12getNbKvHeadsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14GptModelConfig12getNbKvHeadsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14GptModelConfig12getNbKvHeadsEv"></span><span id="tensorrt_llm::runtime::GptModelConfig::getNbKvHeadsCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a51a65e744ce4675cbdaaff83f16f2c54"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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><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_llm7runtime14GptModelConfig12getNbKvHeadsEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig12setNbKvHeadsE8SizeType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig12setNbKvHeadsE8SizeType"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig12setNbKvHeadsE8SizeType"></span><span id="tensorrt_llm::runtime::GptModelConfig::setNbKvHeads__SizeTypeCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1ad837f99f5aafdcc1ef7baa15c86617b5"></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="n"><span class="pre">setNbKvHeads</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">nbKvHeads</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_llm7runtime14GptModelConfig12setNbKvHeadsE8SizeType" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14GptModelConfig13getHiddenSizeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14GptModelConfig13getHiddenSizeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14GptModelConfig13getHiddenSizeEv"></span><span id="tensorrt_llm::runtime::GptModelConfig::getHiddenSizeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a9df3d4f4da4ad6c30e6c91419d6b268f"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><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_llm7runtime14GptModelConfig13getHiddenSizeEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14GptModelConfig14getSizePerHeadEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14GptModelConfig14getSizePerHeadEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14GptModelConfig14getSizePerHeadEv"></span><span id="tensorrt_llm::runtime::GptModelConfig::getSizePerHeadCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a70409542e05a302a44ac2d571aea25db"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><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_llm7runtime14GptModelConfig14getSizePerHeadEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14GptModelConfig11getDataTypeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14GptModelConfig11getDataTypeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14GptModelConfig11getDataTypeEv"></span><span id="tensorrt_llm::runtime::GptModelConfig::getDataTypeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1ab0dd6dd6a8b7cacfe4f60716066f8613"></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_llm7runtime14GptModelConfig11getDataTypeEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14GptModelConfig21useGptAttentionPluginEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14GptModelConfig21useGptAttentionPluginEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14GptModelConfig21useGptAttentionPluginEv"></span><span id="tensorrt_llm::runtime::GptModelConfig::useGptAttentionPluginCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a2eedc76ee7dc4712de143bcdc02f40f9"></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">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_llm7runtime14GptModelConfig21useGptAttentionPluginEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig21useGptAttentionPluginEb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig21useGptAttentionPluginEb"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig21useGptAttentionPluginEb"></span><span id="tensorrt_llm::runtime::GptModelConfig::useGptAttentionPlugin__bCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1ac3723398406d29b02be8a06912ee0d82"></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="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_llm7runtime14GptModelConfig21useGptAttentionPluginEb" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14GptModelConfig14usePackedInputEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14GptModelConfig14usePackedInputEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14GptModelConfig14usePackedInputEv"></span><span id="tensorrt_llm::runtime::GptModelConfig::usePackedInputCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a4894eb60ae226002fda29841977f7935"></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">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_llm7runtime14GptModelConfig14usePackedInputEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig14usePackedInputEb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig14usePackedInputEb"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig14usePackedInputEb"></span><span id="tensorrt_llm::runtime::GptModelConfig::usePackedInput__bCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1af71f4a0b8a8a8fc6e21ea28612b81ed1"></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="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_llm7runtime14GptModelConfig14usePackedInputEb" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14GptModelConfig15usePagedKvCacheEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14GptModelConfig15usePagedKvCacheEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14GptModelConfig15usePagedKvCacheEv"></span><span id="tensorrt_llm::runtime::GptModelConfig::usePagedKvCacheCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a173a85af1a71420f692612ca00c87d69"></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">usePagedKvCache</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_llm7runtime14GptModelConfig15usePagedKvCacheEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig15usePagedKvCacheEb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig15usePagedKvCacheEb"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig15usePagedKvCacheEb"></span><span id="tensorrt_llm::runtime::GptModelConfig::usePagedKvCache__bCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a80165e232642897d7c4fb621bb61fed3"></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="n"><span class="pre">usePagedKvCache</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">pagedKvCache</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_llm7runtime14GptModelConfig15usePagedKvCacheEb" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14GptModelConfig17getTokensPerBlockEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14GptModelConfig17getTokensPerBlockEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14GptModelConfig17getTokensPerBlockEv"></span><span id="tensorrt_llm::runtime::GptModelConfig::getTokensPerBlockCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a0fa158a8e4b14102c6ee79a717d40879"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><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_llm7runtime14GptModelConfig17getTokensPerBlockEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig17setTokensPerBlockE8SizeType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig17setTokensPerBlockE8SizeType"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig17setTokensPerBlockE8SizeType"></span><span id="tensorrt_llm::runtime::GptModelConfig::setTokensPerBlock__SizeTypeCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a443a5a8b94f03dd8b0a8fbbec298d3f8"></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="n"><span class="pre">setTokensPerBlock</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime14GptModelConfig17setTokensPerBlockE8SizeType" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14GptModelConfig12getQuantModeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14GptModelConfig12getQuantModeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14GptModelConfig12getQuantModeEv"></span><span id="tensorrt_llm::runtime::GptModelConfig::getQuantModeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1aa35c989f165bd8aa9437c3c6ba7d0a75"></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">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">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_llm7runtime14GptModelConfig12getQuantModeEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig12setQuantModeEN6common9QuantModeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig12setQuantModeEN6common9QuantModeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig12setQuantModeEN6common9QuantModeE"></span><span id="tensorrt_llm::runtime::GptModelConfig::setQuantMode__common::QuantModeCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a41fe2aa1fdd76ce76301859981425417"></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="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_llm7runtime14GptModelConfig12setQuantModeEN6common9QuantModeE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14GptModelConfig24supportsInflightBatchingEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14GptModelConfig24supportsInflightBatchingEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14GptModelConfig24supportsInflightBatchingEv"></span><span id="tensorrt_llm::runtime::GptModelConfig::supportsInflightBatchingCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a275c4804ce2373f0c977c7a5f2ddd2a0"></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">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_llm7runtime14GptModelConfig24supportsInflightBatchingEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14GptModelConfig15getMaxBatchSizeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14GptModelConfig15getMaxBatchSizeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14GptModelConfig15getMaxBatchSizeEv"></span><span id="tensorrt_llm::runtime::GptModelConfig::getMaxBatchSizeCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a24b66ade6d982663d81fa7b42d9c6055"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><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_llm7runtime14GptModelConfig15getMaxBatchSizeEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig15setMaxBatchSizeE8SizeType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig15setMaxBatchSizeE8SizeType"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig15setMaxBatchSizeE8SizeType"></span><span id="tensorrt_llm::runtime::GptModelConfig::setMaxBatchSize__SizeTypeCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1ab784bd1a00eb729f402abe8875b564ea"></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="n"><span class="pre">setMaxBatchSize</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime14GptModelConfig15setMaxBatchSizeE8SizeType" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14GptModelConfig14getMaxInputLenEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14GptModelConfig14getMaxInputLenEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14GptModelConfig14getMaxInputLenEv"></span><span id="tensorrt_llm::runtime::GptModelConfig::getMaxInputLenCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1af1e370c6c5bbf9dfc854e12c14cccb79"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><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_llm7runtime14GptModelConfig14getMaxInputLenEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig14setMaxInputLenE8SizeType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig14setMaxInputLenE8SizeType"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig14setMaxInputLenE8SizeType"></span><span id="tensorrt_llm::runtime::GptModelConfig::setMaxInputLen__SizeTypeCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1ab80523799abed725a58348c4d56c8ad8"></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="n"><span class="pre">setMaxInputLen</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime14GptModelConfig14setMaxInputLenE8SizeType" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14GptModelConfig15getMaxOutputLenEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14GptModelConfig15getMaxOutputLenEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14GptModelConfig15getMaxOutputLenEv"></span><span id="tensorrt_llm::runtime::GptModelConfig::getMaxOutputLenCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1aaa48e78ace7bb49aa634f3cb84ed7ac4"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">getMaxOutputLen</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_llm7runtime14GptModelConfig15getMaxOutputLenEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig15setMaxOutputLenE8SizeType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig15setMaxOutputLenE8SizeType"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig15setMaxOutputLenE8SizeType"></span><span id="tensorrt_llm::runtime::GptModelConfig::setMaxOutputLen__SizeTypeCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1ab5e91bd0f16db6355ab0e0a14378658a"></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="n"><span class="pre">setMaxOutputLen</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxOutputLen</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_llm7runtime14GptModelConfig15setMaxOutputLenE8SizeType" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14GptModelConfig15getMaxNumTokensEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14GptModelConfig15getMaxNumTokensEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14GptModelConfig15getMaxNumTokensEv"></span><span id="tensorrt_llm::runtime::GptModelConfig::getMaxNumTokensCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a5065aee21fa4e99f5c52bad69117a5cb"></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">optional</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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">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_llm7runtime14GptModelConfig15getMaxNumTokensEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig15setMaxNumTokensENSt8optionalI8SizeTypeEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig15setMaxNumTokensENSt8optionalI8SizeTypeEE"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig15setMaxNumTokensENSt8optionalI8SizeTypeEE"></span><span id="tensorrt_llm::runtime::GptModelConfig::setMaxNumTokens__std::optional:SizeType:CE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a06b325b04c1b396a0430320435f81373"></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="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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime14GptModelConfig15setMaxNumTokensENSt8optionalI8SizeTypeEE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14GptModelConfig20computeContextLogitsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14GptModelConfig20computeContextLogitsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14GptModelConfig20computeContextLogitsEv"></span><span id="tensorrt_llm::runtime::GptModelConfig::computeContextLogitsCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a2f0e4d2c822ec1a83c858674ffce5ec2"></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">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_llm7runtime14GptModelConfig20computeContextLogitsEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig20computeContextLogitsEb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig20computeContextLogitsEb"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig20computeContextLogitsEb"></span><span id="tensorrt_llm::runtime::GptModelConfig::computeContextLogits__bCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a33188908109cfcc7baa824e284a1af8e"></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="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_llm7runtime14GptModelConfig20computeContextLogitsEb" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14GptModelConfig15getModelVariantEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14GptModelConfig15getModelVariantEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14GptModelConfig15getModelVariantEv"></span><span id="tensorrt_llm::runtime::GptModelConfig::getModelVariantC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a9e832f202599100dfac6af143f8c72b4"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14GptModelConfig12ModelVariantE" title="tensorrt_llm::runtime::GptModelConfig::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_llm7runtime14GptModelConfig15getModelVariantEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig15setModelVariantE12ModelVariant">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig15setModelVariantE12ModelVariant"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig15setModelVariantE12ModelVariant"></span><span id="tensorrt_llm::runtime::GptModelConfig::setModelVariant__ModelVariant"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a5fd810a5a863b75656d6c0914b2d5762"></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_llm7runtime14GptModelConfig12ModelVariantE" title="tensorrt_llm::runtime::GptModelConfig::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_llm7runtime14GptModelConfig15setModelVariantE12ModelVariant" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime14GptModelConfig18useCustomAllReduceEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime14GptModelConfig18useCustomAllReduceEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime14GptModelConfig18useCustomAllReduceEv"></span><span id="tensorrt_llm::runtime::GptModelConfig::useCustomAllReduceCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a392c5d57f071a1a36fa628580975faf9"></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">useCustomAllReduce</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_llm7runtime14GptModelConfig18useCustomAllReduceEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig18useCustomAllReduceEb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig18useCustomAllReduceEb"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig18useCustomAllReduceEb"></span><span id="tensorrt_llm::runtime::GptModelConfig::useCustomAllReduce__bCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a79e7448c925f5d6785575250d7a1971f"></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="n"><span class="pre">useCustomAllReduce</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">customAllReduce</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_llm7runtime14GptModelConfig18useCustomAllReduceEb" title="Permalink 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_llm7runtime14GptModelConfig10mVocabSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig10mVocabSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig10mVocabSizeE"></span><span id="tensorrt_llm::runtime::GptModelConfig::mVocabSize__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a18e1092a5c44a311577bc4954efc9ad8"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime14GptModelConfig10mVocabSizeE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig9mNbLayersE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig9mNbLayersE"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig9mNbLayersE"></span><span id="tensorrt_llm::runtime::GptModelConfig::mNbLayers__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a83dd559a542a2ebfe932387d4bc65faa"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime14GptModelConfig9mNbLayersE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig8mNbHeadsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig8mNbHeadsE"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig8mNbHeadsE"></span><span id="tensorrt_llm::runtime::GptModelConfig::mNbHeads__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a3aed4484b10e66d10c0f2e330aea1f06"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime14GptModelConfig8mNbHeadsE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig10mNbKvHeadsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig10mNbKvHeadsE"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig10mNbKvHeadsE"></span><span id="tensorrt_llm::runtime::GptModelConfig::mNbKvHeads__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1aafe290665c7ed003ec2347f04f1a54a5"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mNbKvHeads</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14GptModelConfig10mNbKvHeadsE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig11mHiddenSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig11mHiddenSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig11mHiddenSizeE"></span><span id="tensorrt_llm::runtime::GptModelConfig::mHiddenSize__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a43e64ac4881c3a9eee58b7dee25964e2"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime14GptModelConfig11mHiddenSizeE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig9mDataTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig9mDataTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig9mDataTypeE"></span><span id="tensorrt_llm::runtime::GptModelConfig::mDataType__nvinfer1::DataType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a497a3a345c3e48362790139c6fe56526"></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_llm7runtime14GptModelConfig9mDataTypeE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig22mUseGptAttentionPluginE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig22mUseGptAttentionPluginE"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig22mUseGptAttentionPluginE"></span><span id="tensorrt_llm::runtime::GptModelConfig::mUseGptAttentionPlugin__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a2c837cd8fccbcae0427ccb0334ddd9d2"></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_llm7runtime14GptModelConfig22mUseGptAttentionPluginE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig12mInputPackedE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig12mInputPackedE"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig12mInputPackedE"></span><span id="tensorrt_llm::runtime::GptModelConfig::mInputPacked__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a780f8ca1c1f7565694104b0fa5bf45fb"></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_llm7runtime14GptModelConfig12mInputPackedE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig13mPagedKvCacheE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig13mPagedKvCacheE"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig13mPagedKvCacheE"></span><span id="tensorrt_llm::runtime::GptModelConfig::mPagedKvCache__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a2b70f372d6c64c92f1ad003043dcdb52"></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">mPagedKvCache</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14GptModelConfig13mPagedKvCacheE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig15mTokensPerBlockE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig15mTokensPerBlockE"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig15mTokensPerBlockE"></span><span id="tensorrt_llm::runtime::GptModelConfig::mTokensPerBlock__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1aa5ad1102f45507190c50ccfbd74f0788"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime14GptModelConfig15mTokensPerBlockE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig10mQuantModeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig10mQuantModeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig10mQuantModeE"></span><span id="tensorrt_llm::runtime::GptModelConfig::mQuantMode__common::QuantMode"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a9a5da6ac7818f7df4f68e3306653598b"></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_llm7runtime14GptModelConfig10mQuantModeE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig13mMaxBatchSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig13mMaxBatchSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig13mMaxBatchSizeE"></span><span id="tensorrt_llm::runtime::GptModelConfig::mMaxBatchSize__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a8714ccf704b0a20be40bb99e229c2439"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime14GptModelConfig13mMaxBatchSizeE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig12mMaxInputLenE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig12mMaxInputLenE"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig12mMaxInputLenE"></span><span id="tensorrt_llm::runtime::GptModelConfig::mMaxInputLen__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1ab4be5cfe878d1333c6add3b824b6a7e4"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime14GptModelConfig12mMaxInputLenE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig13mMaxOutputLenE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig13mMaxOutputLenE"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig13mMaxOutputLenE"></span><span id="tensorrt_llm::runtime::GptModelConfig::mMaxOutputLen__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1adc7f72184ca681976df8197f475942c2"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mMaxOutputLen</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14GptModelConfig13mMaxOutputLenE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig13mMaxNumTokensE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig13mMaxNumTokensE"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig13mMaxNumTokensE"></span><span id="tensorrt_llm::runtime::GptModelConfig::mMaxNumTokens__std::optional:SizeType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1aa20bf7021d240c626fa6210511c31c47"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime14GptModelConfig13mMaxNumTokensE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig21mComputeContextLogitsE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig21mComputeContextLogitsE"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig21mComputeContextLogitsE"></span><span id="tensorrt_llm::runtime::GptModelConfig::mComputeContextLogits__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a49736d7c35c69ae8eb50892e69dcfeb3"></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_llm7runtime14GptModelConfig21mComputeContextLogitsE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig13mModelVariantE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig13mModelVariantE"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig13mModelVariantE"></span><span id="tensorrt_llm::runtime::GptModelConfig::mModelVariant__ModelVariant"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1ae45c6e3f4a0ae12a5df7ba02281b544e"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14GptModelConfig12ModelVariantE" title="tensorrt_llm::runtime::GptModelConfig::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_llm7runtime14GptModelConfig13mModelVariantE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14GptModelConfig19mUseCustomAllReduceE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14GptModelConfig19mUseCustomAllReduceE"></span><span id="_CPPv2N12tensorrt_llm7runtime14GptModelConfig19mUseCustomAllReduceE"></span><span id="tensorrt_llm::runtime::GptModelConfig::mUseCustomAllReduce__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptModelConfig_1a2475ff83565a03603f3a3e61425320d6"></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">mUseCustomAllReduce</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14GptModelConfig19mUseCustomAllReduceE" title="Permalink 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="Permalink 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><a class="headerlink" href="#_CPPv4N12tensorrt_llm13batch_managerE" title="Permalink to this definition"></a><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><a class="headerlink" href="#_CPPv4N12tensorrt_llm13batch_manager16kv_cache_managerE" title="Permalink to this definition"></a><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="Permalink 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="Permalink 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_llm7runtime10GptSession10GptSessionERK14GptModelConfigRK11WorldConfigPKvNSt6size_tE9LoggerPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession10GptSessionERK14GptModelConfigRK11WorldConfigPKvNSt6size_tE9LoggerPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession10GptSessionERK14GptModelConfigRK11WorldConfigPKvNSt6size_tE9LoggerPtr"></span><span id="tensorrt_llm::runtime::GptSession::GptSession__GptModelConfigCR.WorldConfigCR.voidCP.std::s.LoggerPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a4b3fd438443e516007a6beb6a27f72dc"></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_llm7runtime14GptModelConfigE" title="tensorrt_llm::runtime::GptModelConfig"><span class="n"><span class="pre">GptModelConfig</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_llm7runtime10GptSession10GptSessionERK14GptModelConfigRK11WorldConfigPKvNSt6size_tE9LoggerPtr" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession10GptSessionERK14GptModelConfigRK11WorldConfigRKNSt6vectorI7uint8_tEE9LoggerPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession10GptSessionERK14GptModelConfigRK11WorldConfigRKNSt6vectorI7uint8_tEE9LoggerPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession10GptSessionERK14GptModelConfigRK11WorldConfigRKNSt6vectorI7uint8_tEE9LoggerPtr"></span><span id="tensorrt_llm::runtime::GptSession::GptSession__GptModelConfigCR.WorldConfigCR.std::vector:uint8_t:CR.LoggerPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a46e03f47ac584b26e20a6429cdacdf18"></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_llm7runtime14GptModelConfigE" title="tensorrt_llm::runtime::GptModelConfig"><span class="n"><span class="pre">GptModelConfig</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_llm7runtime10GptSession10GptSessionERK14GptModelConfigRK11WorldConfigRKNSt6vectorI7uint8_tEE9LoggerPtr" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession10GptSessionERK14GptModelConfigRK11WorldConfigRKNSt6stringE9LoggerPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession10GptSessionERK14GptModelConfigRK11WorldConfigRKNSt6stringE9LoggerPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession10GptSessionERK14GptModelConfigRK11WorldConfigRKNSt6stringE9LoggerPtr"></span><span id="tensorrt_llm::runtime::GptSession::GptSession__GptModelConfigCR.WorldConfigCR.ssCR.LoggerPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a050978a065f4a2b17ec2d0e9ae97a497"></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_llm7runtime14GptModelConfigE" title="tensorrt_llm::runtime::GptModelConfig"><span class="n"><span class="pre">GptModelConfig</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_llm7runtime10GptSession10GptSessionERK14GptModelConfigRK11WorldConfigRKNSt6stringE9LoggerPtr" title="Permalink 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_1a3fa9a3cb9db8934ff06b1df13063b2f3"></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="Permalink 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_1a992c0a12864a9a30ddccc73819fcad62"></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="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="Permalink 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_1aafeb2d6a1276ed948457924d1a54e744"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14GptModelConfigE" title="tensorrt_llm::runtime::GptModelConfig"><span class="n"><span class="pre">GptModelConfig</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="Permalink 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_1a8921d43edbec6aa0b9ba6a6894505542"></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="Permalink 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="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime10GptSession15isCudaGraphModeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime10GptSession15isCudaGraphModeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime10GptSession15isCudaGraphModeEv"></span><span id="tensorrt_llm::runtime::GptSession::isCudaGraphModeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a0aa5b90cb9e19019034d913e326d37b8"></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">isCudaGraphMode</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_llm7runtime10GptSession15isCudaGraphModeEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession16setCudaGraphModeEb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession16setCudaGraphModeEb"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession16setCudaGraphModeEb"></span><span id="tensorrt_llm::runtime::GptSession::setCudaGraphMode__b"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1ac9bbe4bb488194510a54740483878c19"></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">setCudaGraphMode</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">value</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession16setCudaGraphModeEb" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession5setupE8SizeType8SizeType8SizeTypebNSt8optionalI8SizeTypeEENSt8optionalI8SizeTypeEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession5setupE8SizeType8SizeType8SizeTypebNSt8optionalI8SizeTypeEENSt8optionalI8SizeTypeEE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession5setupE8SizeType8SizeType8SizeTypebNSt8optionalI8SizeTypeEENSt8optionalI8SizeTypeEE"></span><span id="tensorrt_llm::runtime::GptSession::setup__SizeType.SizeType.SizeType.b.std::optional:SizeType:.std::optional:SizeType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a3170f16094aafed96e93723675715a16"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxBatchSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxBeamWidth</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxSequenceLength</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>, <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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="n sig-param"><span class="pre">maxTokensInPagedKvCache</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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="n sig-param"><span class="pre">numMicroBatches</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_llm7runtime10GptSession5setupE8SizeType8SizeType8SizeTypebNSt8optionalI8SizeTypeEENSt8optionalI8SizeTypeEE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd><p>Initialize buffers for the given sizes. <code class="docutils literal notranslate"><span class="pre">generate</span></code> may be called with batch size and beam width smaller than the setup parameters. </p>
|
||
<p><code class="docutils literal notranslate"><span class="pre">maxBatchSize</span></code> will be devided by the number of micro batches to initialize each batch buffer. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession8generateER16GenerationOutputRK15GenerationInputRK14SamplingConfig">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession8generateER16GenerationOutputRK15GenerationInputRK14SamplingConfig"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession8generateER16GenerationOutputRK15GenerationInputRK14SamplingConfig"></span><span id="tensorrt_llm::runtime::GptSession::generate__GenerationOutputR.GenerationInputCR.SamplingConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a55d90d8c6fccd7de2975d20ded7e4e42"></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">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="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession8generateER16GenerationOutputRK15GenerationInputRK14SamplingConfig" title="Permalink 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_llm7runtime10GptSession14KvCacheManagerE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession14KvCacheManagerE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession14KvCacheManagerE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a17734c9adff9c0c3dd37d104c9c20879"></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">KvCacheManager</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_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="#_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">KVCacheManager</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession14KvCacheManagerE" title="Permalink 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_llm7runtime10GptSession19generateSingleBatchER16GenerationOutputRK15GenerationInputRK14SamplingConfig">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession19generateSingleBatchER16GenerationOutputRK15GenerationInputRK14SamplingConfig"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession19generateSingleBatchER16GenerationOutputRK15GenerationInputRK14SamplingConfig"></span><span id="tensorrt_llm::runtime::GptSession::generateSingleBatch__GenerationOutputR.GenerationInputCR.SamplingConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a97e3d000bf8d0d49f642228ef659c510"></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">generateSingleBatch</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="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession19generateSingleBatchER16GenerationOutputRK15GenerationInputRK14SamplingConfig" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession18generateMultiBatchER16GenerationOutputRK15GenerationInputRK14SamplingConfig">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession18generateMultiBatchER16GenerationOutputRK15GenerationInputRK14SamplingConfig"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession18generateMultiBatchER16GenerationOutputRK15GenerationInputRK14SamplingConfig"></span><span id="tensorrt_llm::runtime::GptSession::generateMultiBatch__GenerationOutputR.GenerationInputCR.SamplingConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a5ee2425cab580f6fb0d6bd1fb9beedfb"></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">generateMultiBatch</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="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession18generateMultiBatchER16GenerationOutputRK15GenerationInputRK14SamplingConfig" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession14createContextsE8SizeType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession14createContextsE8SizeType"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession14createContextsE8SizeType"></span><span id="tensorrt_llm::runtime::GptSession::createContexts__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a5082e4db217c9ac0c80b0aaf341fdac6"></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><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime10GptSession14createContextsE8SizeType" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession13createBuffersE8SizeType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession13createBuffersE8SizeType"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession13createBuffersE8SizeType"></span><span id="tensorrt_llm::runtime::GptSession::createBuffers__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a9a0d4e7b740eff3848b6cb2d15d99608"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime10GptSession13createBuffersE8SizeType" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession14createDecodersE8SizeType8SizeType8SizeTypeN8nvinfer18DataTypeEb8SizeType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession14createDecodersE8SizeType8SizeType8SizeTypeN8nvinfer18DataTypeEb8SizeType"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession14createDecodersE8SizeType8SizeType8SizeTypeN8nvinfer18DataTypeEb8SizeType"></span><span id="tensorrt_llm::runtime::GptSession::createDecoders__SizeType.SizeType.SizeType.nvinfer1::DataType.b.SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1afeeed7a6643e35e9a917a045fe6e64f3"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">beamWidth</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime10GptSession14createDecodersE8SizeType8SizeType8SizeTypeN8nvinfer18DataTypeEb8SizeType" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession21createKvCacheManagersE8SizeType8SizeType8SizeType8SizeTypeNSt8optionalI8SizeTypeEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession21createKvCacheManagersE8SizeType8SizeType8SizeType8SizeTypeNSt8optionalI8SizeTypeEE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession21createKvCacheManagersE8SizeType8SizeType8SizeType8SizeTypeNSt8optionalI8SizeTypeEE"></span><span id="tensorrt_llm::runtime::GptSession::createKvCacheManagers__SizeType.SizeType.SizeType.SizeType.std::optional:SizeType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a824b4259610fd481825be38792db3326"></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">createKvCacheManagers</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">beamWidth</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxSequenceLength</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">numMicroBatches</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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="n sig-param"><span class="pre">maxTokensInPagedKvCache</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession21createKvCacheManagersE8SizeType8SizeType8SizeType8SizeTypeNSt8optionalI8SizeTypeEE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession30createCustomAllReduceWorkspaceE8SizeType8SizeType8SizeType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession30createCustomAllReduceWorkspaceE8SizeType8SizeType8SizeType"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession30createCustomAllReduceWorkspaceE8SizeType8SizeType8SizeType"></span><span id="tensorrt_llm::runtime::GptSession::createCustomAllReduceWorkspace__SizeType.SizeType.SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a761e1bcaecfcd45d43b9737a97f88ef4"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">beamWidth</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime10GptSession30createCustomAllReduceWorkspaceE8SizeType8SizeType8SizeType" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession16decoderStepAsyncERN7ITensor9SharedPtrERN7ITensor9SharedPtrE8SizeType8SizeType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession16decoderStepAsyncERN7ITensor9SharedPtrERN7ITensor9SharedPtrE8SizeType8SizeType"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession16decoderStepAsyncERN7ITensor9SharedPtrERN7ITensor9SharedPtrE8SizeType8SizeType"></span><span id="tensorrt_llm::runtime::GptSession::decoderStepAsync__ITensor::SharedPtrR.ITensor::SharedPtrR.SizeType.SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1afe77fa5295c0902fa6a6a21a0fac1a02"></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_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="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">outputIds</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="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">newTokens</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">decoderStep</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime10GptSession16decoderStepAsyncERN7ITensor9SharedPtrERN7ITensor9SharedPtrE8SizeType8SizeType" title="Permalink 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_llm7runtime10GptSession14shouldStopSyncE8SizeType8SizeType8SizeType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession14shouldStopSyncE8SizeType8SizeType8SizeType"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession14shouldStopSyncE8SizeType8SizeType8SizeType"></span><span id="tensorrt_llm::runtime::GptSession::shouldStopSync__SizeType.SizeType.SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a4d2ef8994105f56856cd3e44666611e2"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">batchSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">beamWidth</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime10GptSession14shouldStopSyncE8SizeType8SizeType8SizeType" title="Permalink 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_llm7runtime10GptSession17finalizeOutputIdsER7ITensor8SizeType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession17finalizeOutputIdsER7ITensor8SizeType"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession17finalizeOutputIdsER7ITensor8SizeType"></span><span id="tensorrt_llm::runtime::GptSession::finalizeOutputIds__ITensorR.SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1ae8296224f803d653980ffdbddae574e7"></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">finalizeOutputIds</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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime10GptSession17finalizeOutputIdsER7ITensor8SizeType" title="Permalink to this definition"></a><br /></dt>
|
||
<dd><p>Collect final output ids 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_llm7runtime10GptSession19kvCacheAddSequencesE8SizeType8SizeType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession19kvCacheAddSequencesE8SizeType8SizeType"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession19kvCacheAddSequencesE8SizeType8SizeType"></span><span id="tensorrt_llm::runtime::GptSession::kvCacheAddSequences__SizeType.SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1aaf8b7e1bce5d7aafeb3864a5c4bb56ef"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">beamWidth</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime10GptSession19kvCacheAddSequencesE8SizeType8SizeType" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession13initNewTokensERK15GenerationInputRK14SamplingConfig8SizeType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession13initNewTokensERK15GenerationInputRK14SamplingConfig8SizeType"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession13initNewTokensERK15GenerationInputRK14SamplingConfig8SizeType"></span><span id="tensorrt_llm::runtime::GptSession::initNewTokens__GenerationInputCR.SamplingConfigCR.SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a3da5c6695d6822c3e5e585754523c201"></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">initNewTokens</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_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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime10GptSession13initNewTokensERK15GenerationInputRK14SamplingConfig8SizeType" title="Permalink 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__GptModelConfigC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a7c9df79575f744947edb70aabeceb1ae"></span><span class="k"><span class="pre">const</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14GptModelConfigE" title="tensorrt_llm::runtime::GptModelConfig"><span class="n"><span class="pre">GptModelConfig</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_llm7runtime10GptSession12mModelConfigE" title="Permalink 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_1a718d9218f4e18056a985e79d6d301747"></span><span class="k"><span class="pre">const</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">mWorldConfig</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession12mWorldConfigE" title="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a3eb70f53bb0cbc5c6e032c37520b101c"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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="Permalink 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="Permalink 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="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession16mNumMicroBatchesE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession16mNumMicroBatchesE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession16mNumMicroBatchesE"></span><span id="tensorrt_llm::runtime::GptSession::mNumMicroBatches__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a25c4baa3f042dae7e97ea316a746a538"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mNumMicroBatches</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession16mNumMicroBatchesE" title="Permalink 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="Permalink 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="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession16mKvCacheManagersE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession16mKvCacheManagersE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession16mKvCacheManagersE"></span><span id="tensorrt_llm::runtime::GptSession::mKvCacheManagers__std::vector:std::shared_ptr:KvCacheManager::"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a09318539e6548be6001a61d0fecb69fa"></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_llm7runtime10GptSession14KvCacheManagerE" title="tensorrt_llm::runtime::GptSession::KvCacheManager"><span class="n"><span class="pre">KvCacheManager</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">mKvCacheManagers</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession16mKvCacheManagersE" title="Permalink 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="Permalink 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="Permalink 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::array:CudaGraphExecutor.2:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1a3c80743c41fb287fbc0eb5723bbefa7b"></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="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="m"><span class="pre">2</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">mCudaGraphInstances</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession19mCudaGraphInstancesE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor16prepareNextGraphERK11TllmRuntime8SizeType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor16prepareNextGraphERK11TllmRuntime8SizeType"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor16prepareNextGraphERK11TllmRuntime8SizeType"></span><span id="tensorrt_llm::runtime::GptSession::CudaGraphExecutor::prepareNextGraph__TllmRuntimeCR.SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1CudaGraphExecutor_1ae8bcb9ec8982af90576b5fc24b6f5599"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime10GptSession17CudaGraphExecutor16prepareNextGraphERK11TllmRuntime8SizeType" title="Permalink 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="Permalink 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_llm7runtime10GptSession17CudaGraphExecutor16cudaGraphExecPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor16cudaGraphExecPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor16cudaGraphExecPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1CudaGraphExecutor_1a6e7ee051d29b4a35c5ecb280404cef77"></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">cudaGraphExecPtr</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">cudaGraphExec_t</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor16cudaGraphExecPtrE" title="Permalink 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="Permalink 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="Permalink 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="Permalink 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__cudaGraphExecPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1GptSession_1_1CudaGraphExecutor_1afb0fac5805c84a909296f85363c505ed"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime10GptSession17CudaGraphExecutor16cudaGraphExecPtrE" title="tensorrt_llm::runtime::GptSession::CudaGraphExecutor::cudaGraphExecPtr"><span class="n"><span class="pre">cudaGraphExecPtr</span></span></a><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="Permalink 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="Permalink 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_1ae397b7fd72675fbc97c0e662cc71b5eb"></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="Permalink 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="Permalink 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="Permalink 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_1adc272a863661fdf8ea08ad956ebe5975"></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="s"><span class="pre">"GPU"</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kGPUEE5valueE" title="Permalink 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="Permalink 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_1ad18d7d70b64c48821a5928c3eb1b6d0e"></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="s"><span class="pre">"CPU"</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType4kCPUEE5valueE" title="Permalink 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="Permalink 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_1a661ffce25b1fb3d0f97348ef5e92e6b6"></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="s"><span class="pre">"PINNED"</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime16MemoryTypeStringIN10MemoryType7kPINNEDEE5valueE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kFLOATEEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kFLOATEEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kFLOATEEE"></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_1CppDataType_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">CppDataType</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_llm7runtime11CppDataTypeIN8nvinfer18DataType6kFLOATEEE" title="Permalink 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_llm7runtime11CppDataTypeIN8nvinfer18DataType6kFLOATEE4typeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kFLOATEE4typeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kFLOATEE4typeE"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1CppDataType_3_01nvinfer1_1_1DataType_1_1kFLOAT_01_4_1aae52e3ccf7e29b666cd03fec6539e026"></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_llm7runtime11CppDataTypeIN8nvinfer18DataType6kFLOATEE4typeE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType5kHALFEEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType5kHALFEEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType5kHALFEEE"></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_1CppDataType_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">CppDataType</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_llm7runtime11CppDataTypeIN8nvinfer18DataType5kHALFEEE" title="Permalink 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_llm7runtime11CppDataTypeIN8nvinfer18DataType5kHALFEE4typeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType5kHALFEE4typeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType5kHALFEE4typeE"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1CppDataType_3_01nvinfer1_1_1DataType_1_1kHALF_01_4_1adb4c7442a60ad648dc57760e15affff4"></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_llm7runtime11CppDataTypeIN8nvinfer18DataType5kHALFEE4typeE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType5kINT8EEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType5kINT8EEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType5kINT8EEE"></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_1CppDataType_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">CppDataType</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_llm7runtime11CppDataTypeIN8nvinfer18DataType5kINT8EEE" title="Permalink 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_llm7runtime11CppDataTypeIN8nvinfer18DataType5kINT8EE4typeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType5kINT8EE4typeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType5kINT8EE4typeE"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1CppDataType_3_01nvinfer1_1_1DataType_1_1kINT8_01_4_1aaa4acb1449b214520aa455642c7d95c8"></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_llm7runtime11CppDataTypeIN8nvinfer18DataType5kINT8EE4typeE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT32EEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT32EEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT32EEE"></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_1CppDataType_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">CppDataType</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_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT32EEE" title="Permalink 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_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT32EE4typeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT32EE4typeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT32EE4typeE"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1CppDataType_3_01nvinfer1_1_1DataType_1_1kINT32_01_4_1a4f40534926f2d5f89e2dfac85ff27c2b"></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_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT32EE4typeE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT64EEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT64EEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT64EEE"></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_1CppDataType_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">CppDataType</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_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT64EEE" title="Permalink 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_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT64EE4typeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT64EE4typeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT64EE4typeE"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1CppDataType_3_01nvinfer1_1_1DataType_1_1kINT64_01_4_1a7d116567f78612c70c940df5fe74f981"></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_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT64EE4typeE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT32EXL1EEEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT32EXL1EEEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT32EXtrueEEE"></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_1CppDataType_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">CppDataType</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_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT32EXL1EEEE" title="Permalink 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_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT32EXL1EEE4typeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT32EXL1EEE4typeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT32EXtrueEE4typeE"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1CppDataType_3_01nvinfer1_1_1DataType_1_1kINT32_00_01true_01_4_1afc8c6e195c4db3b75d5bbe02edc8142e"></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_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT32EXL1EEE4typeE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp" id="_CPPv4IEN12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT64EXL1EEEE">
|
||
<span id="_CPPv3IEN12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT64EXL1EEEE"></span><span id="_CPPv2IEN12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT64EXtrueEEE"></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_1CppDataType_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">CppDataType</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_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT64EXL1EEEE" title="Permalink 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_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT64EXL1EEE4typeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT64EXL1EEE4typeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT64EXtrueEE4typeE"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1CppDataType_3_01nvinfer1_1_1DataType_1_1kINT64_00_01true_01_4_1a5963e497e488b6617c842048bcb4a07c"></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_llm7runtime11CppDataTypeIN8nvinfer18DataType6kINT64EXL1EEE4typeE" title="Permalink 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_llm7runtime11CppDataTypeIN8nvinfer18DataType5kBOOLE9kUnsignedEE">
|
||
<span id="_CPPv3I_bEN12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType5kBOOLE9kUnsignedEE"></span><span id="_CPPv2I_bEN12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType5kBOOLE9kUnsignedEE"></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_1CppDataType_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">CppDataType</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_llm7runtime11CppDataTypeIN8nvinfer18DataType5kBOOLE9kUnsignedEE" title="tensorrt_llm::runtime::CppDataType<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_llm7runtime11CppDataTypeIN8nvinfer18DataType5kBOOLE9kUnsignedEE" title="Permalink 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_llm7runtime11CppDataTypeIN8nvinfer18DataType5kBOOLE9kUnsignedE4typeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType5kBOOLE9kUnsignedE4typeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType5kBOOLE9kUnsignedE4typeE"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1CppDataType_3_01nvinfer1_1_1DataType_1_1kBOOL_00_01kUnsigned_01_4_1aabba6c53746defcf62a8448c7511cdcb"></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_llm7runtime11CppDataTypeIN8nvinfer18DataType5kBOOLE9kUnsignedE4typeE" title="Permalink 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_llm7runtime11CppDataTypeIN8nvinfer18DataType6kUINT8E9kUnsignedEE">
|
||
<span id="_CPPv3I_bEN12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kUINT8E9kUnsignedEE"></span><span id="_CPPv2I_bEN12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kUINT8E9kUnsignedEE"></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_1CppDataType_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">CppDataType</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_llm7runtime11CppDataTypeIN8nvinfer18DataType6kUINT8E9kUnsignedEE" title="tensorrt_llm::runtime::CppDataType<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_llm7runtime11CppDataTypeIN8nvinfer18DataType6kUINT8E9kUnsignedEE" title="Permalink 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_llm7runtime11CppDataTypeIN8nvinfer18DataType6kUINT8E9kUnsignedE4typeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kUINT8E9kUnsignedE4typeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11CppDataTypeIN8nvinfer18DataType6kUINT8E9kUnsignedE4typeE"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1CppDataType_3_01nvinfer1_1_1DataType_1_1kUINT8_00_01kUnsigned_01_4_1ad9bc02ed66999f1a5f09850f36dfc8ad"></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_llm7runtime11CppDataTypeIN8nvinfer18DataType6kUINT8E9kUnsignedE4typeE" title="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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_1ad7234909a7d00791f3d22a797078e9de"></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="Permalink 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_1a1398d57cec6aacd5be041575ccb140fb"></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="Permalink 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_1aa39a24d2db1c10a205c75aa2c6a8f069"></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="Permalink to this definition"></a><br /></dt>
|
||
<dd></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_1a2f29317ef105d176296ef6b5518bcddf"></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="Permalink to this definition"></a><br /></dt>
|
||
<dd></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_1a4363785cd968e4e0cd1933cba37850b9"></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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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_1a10497366a0660dd533bd6bac96a65e97"></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">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="Permalink 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="Permalink 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="Permalink 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="Permalink 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><a class="headerlink" href="#_CPPv4I0EN12tensorrt_llm7runtime11BufferRangeE" title="Permalink 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_llm7runtime11BufferRange10value_typeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11BufferRange10value_typeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11BufferRange10value_typeE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferRange_1a577abddadeb9927120873f9d07325e37"></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">value_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="#_CPPv4I0EN12tensorrt_llm7runtime11BufferRangeE" title="tensorrt_llm::runtime::BufferRange::T"><span class="n"><span class="pre">T</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange10value_typeE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11BufferRange9size_typeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11BufferRange9size_typeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11BufferRange9size_typeE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferRange_1ac52032549a72c5078e9dea0d731626ab"></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">size_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">size_t</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange9size_typeE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11BufferRange9referenceE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11BufferRange9referenceE"></span><span id="_CPPv2N12tensorrt_llm7runtime11BufferRange9referenceE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferRange_1acb95456868b359618379a83adc048eaf"></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">reference</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_llm7runtime11BufferRange10value_typeE" title="tensorrt_llm::runtime::BufferRange::value_type"><span class="n"><span class="pre">value_type</span></span></a><span class="p"><span class="pre">&</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange9referenceE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11BufferRange15const_referenceE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11BufferRange15const_referenceE"></span><span id="_CPPv2N12tensorrt_llm7runtime11BufferRange15const_referenceE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferRange_1aed057be953bd7fb90a48f5a7c3ae1692"></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">const_reference</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_llm7runtime11BufferRange10value_typeE" title="tensorrt_llm::runtime::BufferRange::value_type"><span class="n"><span class="pre">value_type</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_llm7runtime11BufferRange15const_referenceE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11BufferRange7pointerE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11BufferRange7pointerE"></span><span id="_CPPv2N12tensorrt_llm7runtime11BufferRange7pointerE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferRange_1a286cbcbcb22a4f81104c9124130118c3"></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><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_llm7runtime11BufferRange7pointerE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11BufferRange13const_pointerE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11BufferRange13const_pointerE"></span><span id="_CPPv2N12tensorrt_llm7runtime11BufferRange13const_pointerE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferRange_1aad04e26d81c56579e3d6a69ac5fa7412"></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">const_pointer</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="w"> </span><span class="k"><span class="pre">const</span></span><span class="p"><span class="pre">*</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange13const_pointerE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11BufferRange8iteratorE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11BufferRange8iteratorE"></span><span id="_CPPv2N12tensorrt_llm7runtime11BufferRange8iteratorE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferRange_1aeb69cf6a643d84e5c7d051bee4311869"></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">iterator</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_llm7runtime11BufferRange7pointerE" title="tensorrt_llm::runtime::BufferRange::pointer"><span class="n"><span class="pre">pointer</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange8iteratorE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11BufferRange14const_iteratorE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11BufferRange14const_iteratorE"></span><span id="_CPPv2N12tensorrt_llm7runtime11BufferRange14const_iteratorE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferRange_1a36e1f97c5d72f04e004f1624a5709986"></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">const_iterator</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_llm7runtime11BufferRange13const_pointerE" title="tensorrt_llm::runtime::BufferRange::const_pointer"><span class="n"><span class="pre">const_pointer</span></span></a><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange14const_iteratorE" title="Permalink 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_llm7runtime11BufferRange11BufferRangeER7IBuffer">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11BufferRange11BufferRangeER7IBuffer"></span><span id="_CPPv2N12tensorrt_llm7runtime11BufferRange11BufferRangeER7IBuffer"></span><span id="tensorrt_llm::runtime::BufferRange::BufferRange__IBufferR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferRange_1a28e8d9ed5250f967aa8f06317074af7b"></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="#_CPPv4N12tensorrt_llm7runtime11BufferRange11BufferRangeER7IBuffer" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11BufferRange5beginEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11BufferRange5beginEv"></span><span id="_CPPv2N12tensorrt_llm7runtime11BufferRange5beginEv"></span><span id="tensorrt_llm::runtime::BufferRange::begin"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferRange_1aadf490250d5ce134a93e32ee11a34980"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange8iteratorE" title="tensorrt_llm::runtime::BufferRange::iterator"><span class="n"><span class="pre">iterator</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">begin</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange5beginEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11BufferRange3endEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11BufferRange3endEv"></span><span id="_CPPv2N12tensorrt_llm7runtime11BufferRange3endEv"></span><span id="tensorrt_llm::runtime::BufferRange::end"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferRange_1aefcb27b14d94348a6fad611d64fbac92"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange8iteratorE" title="tensorrt_llm::runtime::BufferRange::iterator"><span class="n"><span class="pre">iterator</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">end</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange3endEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11BufferRange5beginEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11BufferRange5beginEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11BufferRange5beginEv"></span><span id="tensorrt_llm::runtime::BufferRange::beginC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferRange_1a7fd5ff6c02c395fa4237a0eb1d9330cf"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange14const_iteratorE" title="tensorrt_llm::runtime::BufferRange::const_iterator"><span class="n"><span class="pre">const_iterator</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">begin</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_llm7runtime11BufferRange5beginEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11BufferRange3endEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11BufferRange3endEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11BufferRange3endEv"></span><span id="tensorrt_llm::runtime::BufferRange::endC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferRange_1a758833712fb9f5aac0ec22591d767fee"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange14const_iteratorE" title="tensorrt_llm::runtime::BufferRange::const_iterator"><span class="n"><span class="pre">const_iterator</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">end</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_llm7runtime11BufferRange3endEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11BufferRange6cbeginEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11BufferRange6cbeginEv"></span><span id="_CPPv2N12tensorrt_llm7runtime11BufferRange6cbeginEv"></span><span id="tensorrt_llm::runtime::BufferRange::cbegin"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferRange_1aad76274f379a648c172fa9f064f88d62"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange14const_iteratorE" title="tensorrt_llm::runtime::BufferRange::const_iterator"><span class="n"><span class="pre">const_iterator</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">cbegin</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange6cbeginEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11BufferRange4cendEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11BufferRange4cendEv"></span><span id="_CPPv2N12tensorrt_llm7runtime11BufferRange4cendEv"></span><span id="tensorrt_llm::runtime::BufferRange::cend"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferRange_1a28a237e06e8971046f0fd4d7d9011fff"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange14const_iteratorE" title="tensorrt_llm::runtime::BufferRange::const_iterator"><span class="n"><span class="pre">const_iterator</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">cend</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange4cendEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11BufferRange6cbeginEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11BufferRange6cbeginEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11BufferRange6cbeginEv"></span><span id="tensorrt_llm::runtime::BufferRange::cbeginC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferRange_1a8e5dfa0b4e150b5fdc0ce78af3c5d243"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange14const_iteratorE" title="tensorrt_llm::runtime::BufferRange::const_iterator"><span class="n"><span class="pre">const_iterator</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">cbegin</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_llm7runtime11BufferRange6cbeginEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11BufferRange4cendEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11BufferRange4cendEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11BufferRange4cendEv"></span><span id="tensorrt_llm::runtime::BufferRange::cendC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferRange_1a5df247c962064585ba1c47242a562dea"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange14const_iteratorE" title="tensorrt_llm::runtime::BufferRange::const_iterator"><span class="n"><span class="pre">const_iterator</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">cend</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_llm7runtime11BufferRange4cendEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11BufferRange4sizeEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11BufferRange4sizeEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime11BufferRange4sizeEv"></span><span id="tensorrt_llm::runtime::BufferRange::sizeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferRange_1add27991a49b2ff4b17b187711cd9eba6"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange9size_typeE" title="tensorrt_llm::runtime::BufferRange::size_type"><span class="n"><span class="pre">size_type</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">size</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_llm7runtime11BufferRange4sizeEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11BufferRangeixE9size_type">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11BufferRangeixE9size_type"></span><span id="_CPPv2N12tensorrt_llm7runtime11BufferRangeixE9size_type"></span><span id="tensorrt_llm::runtime::BufferRange::subscript-operator__size_type"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferRange_1ab65060237344abbc8db90f2062abb3ab"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange9referenceE" title="tensorrt_llm::runtime::BufferRange::reference"><span class="n"><span class="pre">reference</span></span></a><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_llm7runtime11BufferRange9size_typeE" title="tensorrt_llm::runtime::BufferRange::size_type"><span class="n"><span class="pre">size_type</span></span></a><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_llm7runtime11BufferRangeixE9size_type" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime11BufferRangeixE9size_type">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime11BufferRangeixE9size_type"></span><span id="_CPPv2NK12tensorrt_llm7runtime11BufferRangeixE9size_type"></span><span id="tensorrt_llm::runtime::BufferRange::subscript-operator__size_typeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferRange_1acc8eff6f7541d117de3e09730425a430"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange15const_referenceE" title="tensorrt_llm::runtime::BufferRange::const_reference"><span class="n"><span class="pre">const_reference</span></span></a><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_llm7runtime11BufferRange9size_typeE" title="tensorrt_llm::runtime::BufferRange::size_type"><span class="n"><span class="pre">size_type</span></span></a><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_llm7runtime11BufferRangeixE9size_type" title="Permalink 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_llm7runtime11BufferRange5mDataE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11BufferRange5mDataE"></span><span id="_CPPv2N12tensorrt_llm7runtime11BufferRange5mDataE"></span><span id="tensorrt_llm::runtime::BufferRange::mData__TP"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferRange_1afd541e9707bd79322c5290261787902e"></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="sig-name descname"><span class="n"><span class="pre">mData</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange5mDataE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11BufferRange5mSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11BufferRange5mSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11BufferRange5mSizeE"></span><span id="tensorrt_llm::runtime::BufferRange::mSize__size_type"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1BufferRange_1ae18e3afe46fa34b6f3a8d5bfa7911bfc"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange9size_typeE" title="tensorrt_llm::runtime::BufferRange::size_type"><span class="n"><span class="pre">size_type</span></span></a><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">mSize</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11BufferRange5mSizeE" title="Permalink 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_llm7runtime11CppDataTypeE">
|
||
<span id="_CPPv3I_N8nvinfer18DataTypeE_b_bEN12tensorrt_llm7runtime11CppDataTypeE"></span><span id="_CPPv2I_N8nvinfer18DataTypeE_b_bEN12tensorrt_llm7runtime11CppDataTypeE"></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_1CppDataType"></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">CppDataType</span></span></span><a class="headerlink" href="#_CPPv4I_N8nvinfer18DataTypeE_b_bEN12tensorrt_llm7runtime11CppDataTypeE" title="Permalink 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_llm7runtime11CppDataTypeI9kDataType9kUnsignedXL1EEEE">
|
||
<span id="_CPPv3I_N8nvinfer18DataTypeE_bEN12tensorrt_llm7runtime11CppDataTypeI9kDataType9kUnsignedXL1EEEE"></span><span id="_CPPv2I_N8nvinfer18DataTypeE_bEN12tensorrt_llm7runtime11CppDataTypeI9kDataType9kUnsignedXtrueEEE"></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_1CppDataType_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">CppDataType</span></span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I_N8nvinfer18DataTypeE_bEN12tensorrt_llm7runtime11CppDataTypeI9kDataType9kUnsignedXL1EEEE" title="tensorrt_llm::runtime::CppDataType<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_llm7runtime11CppDataTypeI9kDataType9kUnsignedXL1EEEE" title="tensorrt_llm::runtime::CppDataType<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_llm7runtime11CppDataTypeI9kDataType9kUnsignedXL1EEEE" title="Permalink 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_llm7runtime11CppDataTypeI9kDataType9kUnsignedXL1EEE4typeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11CppDataTypeI9kDataType9kUnsignedXL1EEE4typeE"></span><span id="_CPPv2N12tensorrt_llm7runtime11CppDataTypeI9kDataType9kUnsignedXtrueEE4typeE"></span><span class="target" id="structtensorrt__llm_1_1runtime_1_1CppDataType_3_01kDataType_00_01kUnsigned_00_01true_01_4_1a158884a279cc456ecfef4acad4f260a8"></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">CppDataType</span></span><span class="p"><span class="pre"><</span></span><a class="reference internal" href="#_CPPv4I_N8nvinfer18DataTypeE_bEN12tensorrt_llm7runtime11CppDataTypeI9kDataType9kUnsignedXL1EEEE" title="tensorrt_llm::runtime::CppDataType<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_llm7runtime11CppDataTypeI9kDataType9kUnsignedXL1EEEE" title="tensorrt_llm::runtime::CppDataType<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_llm7runtime11CppDataTypeI9kDataType9kUnsignedXL1EEE4typeE" title="Permalink 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_1CppDataType_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_1CppDataType_3_01nvinfer1_1_1DataType_1_1kBOOL_00_01kUnsigned_01_4_1aabba6c53746defcf62a8448c7511cdcb"></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="Permalink 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_1CppDataType_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_1CppDataType_3_01nvinfer1_1_1DataType_1_1kFLOAT_01_4_1aae52e3ccf7e29b666cd03fec6539e026"></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>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1CppDataType_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_1CppDataType_3_01nvinfer1_1_1DataType_1_1kHALF_01_4_1adb4c7442a60ad648dc57760e15affff4"></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>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1CppDataType_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_1CppDataType_3_01nvinfer1_1_1DataType_1_1kINT32_01_4_1a4f40534926f2d5f89e2dfac85ff27c2b"></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>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1CppDataType_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_1CppDataType_3_01nvinfer1_1_1DataType_1_1kINT32_00_01true_01_4_1afc8c6e195c4db3b75d5bbe02edc8142e"></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>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1CppDataType_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_1CppDataType_3_01nvinfer1_1_1DataType_1_1kINT64_01_4_1a7d116567f78612c70c940df5fe74f981"></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>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1CppDataType_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_1CppDataType_3_01nvinfer1_1_1DataType_1_1kINT64_00_01true_01_4_1a5963e497e488b6617c842048bcb4a07c"></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>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1CppDataType_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_1CppDataType_3_01nvinfer1_1_1DataType_1_1kINT8_01_4_1aaa4acb1449b214520aa455642c7d95c8"></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>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp struct">
|
||
<dt class="sig sig-object cpp">
|
||
<span class="target" id="structtensorrt__llm_1_1runtime_1_1CppDataType_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_1CppDataType_3_01nvinfer1_1_1DataType_1_1kUINT8_00_01kUnsigned_01_4_1ad9bc02ed66999f1a5f09850f36dfc8ad"></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>
|
||
</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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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_1a1b29144573f04c11a8c59576a2ddca4f"></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="Permalink 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_1a49f4130965beadb2b82a0b6bcd50cd11"></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="Permalink 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_1ada0b7259447dab20f65e144030f55cef"></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="Permalink 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_1a797f41a093bb00596edaf60f329f66e1"></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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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_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="Permalink 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="_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="Permalink 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="Permalink 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="Permalink 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="Permalink 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_1a10744c55831bf12db698adbb511d660e"></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="Permalink 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_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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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_1ad18d7d70b64c48821a5928c3eb1b6d0e"></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="s"><span class="pre">"CPU"</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime19PhonyNameDueToError5valueE" title="Permalink 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_1adc272a863661fdf8ea08ad956ebe5975"></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="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_1a661ffce25b1fb3d0f97348ef5e92e6b6"></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="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" 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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_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="Permalink 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_1ae513ee17ebefd7e2c538476736488aa1"></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_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="Permalink 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_1aafb13ea78bfe5c1ddf1a77093531b0ff"></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">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><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="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="Permalink 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="Permalink 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="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="igptdecoderbatch-h">
|
||
<h2>iGptDecoderBatch.h<a class="headerlink" href="#igptdecoderbatch-h" title="Permalink 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_llm7runtime16IGptDecoderBatchE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16IGptDecoderBatchE"></span><span id="_CPPv2N12tensorrt_llm7runtime16IGptDecoderBatchE"></span><span id="tensorrt_llm::runtime::IGptDecoderBatch"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatch"></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">IGptDecoderBatch</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="#_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_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_llm7runtime16IGptDecoderBatchE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd><div class="docutils container">
|
||
<em>#include <iGptDecoderBatch.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_1GptDecoderBatch"><span class="std std-ref">tensorrt_llm::runtime::GptDecoderBatch</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_llm7runtime16IGptDecoderBatch13CudaStreamPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16IGptDecoderBatch13CudaStreamPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime16IGptDecoderBatch13CudaStreamPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatch_1a8968dd741ba1d28cddca68d850a62f59"></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_llm7runtime16IGptDecoderBatch13CudaStreamPtrE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime16IGptDecoderBatch9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16IGptDecoderBatch9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime16IGptDecoderBatch9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatch_1ac23bb332c54e6c2b665d5f35fc64baf8"></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_llm7runtime16IGptDecoderBatch9TensorPtrE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp type">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime16IGptDecoderBatch8TokenPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16IGptDecoderBatch8TokenPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime16IGptDecoderBatch8TokenPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatch_1aefe802b00d9cc338c76f2bfc7bafad4c"></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">TokenPtr</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_batch5TokenE" title="tensorrt_llm::runtime::decoder_batch::Token"><span class="n"><span class="pre">Token</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_llm7runtime16IGptDecoderBatch8TokenPtrE" title="Permalink 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_llm7runtime16IGptDecoderBatch10newRequestE8SizeTypeRKN13decoder_batch7RequestERK14SamplingConfig">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16IGptDecoderBatch10newRequestE8SizeTypeRKN13decoder_batch7RequestERK14SamplingConfig"></span><span id="_CPPv2N12tensorrt_llm7runtime16IGptDecoderBatch10newRequestE8SizeTypeRKN13decoder_batch7RequestERK14SamplingConfig"></span><span id="tensorrt_llm::runtime::IGptDecoderBatch::newRequest__SizeType.decoder_batch::RequestCR.SamplingConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatch_1a795d229a59bc66688d9609ce62b40775"></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">newRequest</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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><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_llm7runtime16IGptDecoderBatch10newRequestE8SizeTypeRKN13decoder_batch7RequestERK14SamplingConfig" title="Permalink to this definition"></a><br /></dt>
|
||
<dd><p>Initialize the decoder at <code class="docutils literal notranslate"><span class="pre">batchIdx</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_llm7runtime16IGptDecoderBatch12forwardAsyncERN13decoder_batch6OutputERKN13decoder_batch5InputE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16IGptDecoderBatch12forwardAsyncERN13decoder_batch6OutputERKN13decoder_batch5InputE"></span><span id="_CPPv2N12tensorrt_llm7runtime16IGptDecoderBatch12forwardAsyncERN13decoder_batch6OutputERKN13decoder_batch5InputE"></span><span id="tensorrt_llm::runtime::IGptDecoderBatch::forwardAsync__decoder_batch::OutputR.decoder_batch::InputCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatch_1a6c990b4aa463031ee4d4e2724997ae0e"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16IGptDecoderBatch8TokenPtrE" title="tensorrt_llm::runtime::IGptDecoderBatch::TokenPtr"><span class="n"><span class="pre">TokenPtr</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_llm7runtime16IGptDecoderBatch12forwardAsyncERN13decoder_batch6OutputERKN13decoder_batch5InputE" title="Permalink 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_llm7runtime16IGptDecoderBatch11forwardSyncERKN13decoder_batch5TokenE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16IGptDecoderBatch11forwardSyncERKN13decoder_batch5TokenE"></span><span id="_CPPv2N12tensorrt_llm7runtime16IGptDecoderBatch11forwardSyncERKN13decoder_batch5TokenE"></span><span id="tensorrt_llm::runtime::IGptDecoderBatch::forwardSync__decoder_batch::TokenCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatch_1a7171eb7513202c09360d4139117c424e"></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_batch5TokenE" title="tensorrt_llm::runtime::decoder_batch::Token"><span class="n"><span class="pre">Token</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_llm7runtime16IGptDecoderBatch11forwardSyncERKN13decoder_batch5TokenE" title="Permalink 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_llm7runtime16IGptDecoderBatch7forwardERN13decoder_batch6OutputERKN13decoder_batch5InputE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16IGptDecoderBatch7forwardERN13decoder_batch6OutputERKN13decoder_batch5InputE"></span><span id="_CPPv2N12tensorrt_llm7runtime16IGptDecoderBatch7forwardERN13decoder_batch6OutputERKN13decoder_batch5InputE"></span><span id="tensorrt_llm::runtime::IGptDecoderBatch::forward__decoder_batch::OutputR.decoder_batch::InputCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatch_1ab7e07276f33bfa6ef3c90baa49ee6eec"></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_llm7runtime16IGptDecoderBatch7forwardERN13decoder_batch6OutputERKN13decoder_batch5InputE" title="Permalink 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_llm7runtime16IGptDecoderBatch12getOutputIdsE8SizeType">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime16IGptDecoderBatch12getOutputIdsE8SizeType"></span><span id="_CPPv2NK12tensorrt_llm7runtime16IGptDecoderBatch12getOutputIdsE8SizeType"></span><span id="tensorrt_llm::runtime::IGptDecoderBatch::getOutputIds__SizeTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatch_1a56bbbba5d310262f3614c9ca16d6e1d7"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16IGptDecoderBatch9TensorPtrE" title="tensorrt_llm::runtime::IGptDecoderBatch::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">getOutputIds</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime16IGptDecoderBatch12getOutputIdsE8SizeType" title="Permalink 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, 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_llm7runtime16IGptDecoderBatch17getFinalOutputIdsE8SizeType">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime16IGptDecoderBatch17getFinalOutputIdsE8SizeType"></span><span id="_CPPv2NK12tensorrt_llm7runtime16IGptDecoderBatch17getFinalOutputIdsE8SizeType"></span><span id="tensorrt_llm::runtime::IGptDecoderBatch::getFinalOutputIds__SizeTypeC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatch_1a5b4e85c580d52fed45a727a4baf0ab0e"></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">tuple</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><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16IGptDecoderBatch9TensorPtrE" title="tensorrt_llm::runtime::IGptDecoderBatch::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">getFinalOutputIds</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime16IGptDecoderBatch17getFinalOutputIdsE8SizeType" title="Permalink to this definition"></a><br /></dt>
|
||
<dd><p>Execute postProcessRequest and returns OutputIds for request <code class="docutils literal notranslate"><span class="pre">batchIdx</span></code>. Result will only be available after event returned </p>
|
||
<dl class="field-list simple">
|
||
<dt class="field-odd">Returns<span class="colon">:</span></dt>
|
||
<dd class="field-odd"><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_llm7runtime16IGptDecoderBatch16getFinishedBeamsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime16IGptDecoderBatch16getFinishedBeamsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime16IGptDecoderBatch16getFinishedBeamsEv"></span><span id="tensorrt_llm::runtime::IGptDecoderBatch::getFinishedBeamsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatch_1ae3ea5080726f1ce7b14961eca6827e25"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16IGptDecoderBatch9TensorPtrE" title="tensorrt_llm::runtime::IGptDecoderBatch::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">getFinishedBeams</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_llm7runtime16IGptDecoderBatch16getFinishedBeamsEv" title="Permalink 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], marks finished requests (per beam), on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime16IGptDecoderBatch16getOutputLengthsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime16IGptDecoderBatch16getOutputLengthsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime16IGptDecoderBatch16getOutputLengthsEv"></span><span id="tensorrt_llm::runtime::IGptDecoderBatch::getOutputLengthsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatch_1a8f082a6761daadd109c357f30c501433"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16IGptDecoderBatch9TensorPtrE" title="tensorrt_llm::runtime::IGptDecoderBatch::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">getOutputLengths</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_llm7runtime16IGptDecoderBatch16getOutputLengthsEv" title="Permalink 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], total sequence lengths (per beam), on gpu </p>
|
||
</dd>
|
||
</dl>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime16IGptDecoderBatch11getFinishedEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime16IGptDecoderBatch11getFinishedEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime16IGptDecoderBatch11getFinishedEv"></span><span id="tensorrt_llm::runtime::IGptDecoderBatch::getFinishedC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatch_1aaa0b5e6c02b7043c3b29a1d8ca68df76"></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_llm7runtime16IGptDecoderBatch11getFinishedEv" title="Permalink 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_llm7runtime16IGptDecoderBatch14getCumLogProbsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime16IGptDecoderBatch14getCumLogProbsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime16IGptDecoderBatch14getCumLogProbsEv"></span><span id="tensorrt_llm::runtime::IGptDecoderBatch::getCumLogProbsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatch_1afa130a3df00b7d86e280f5caf017ab9d"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16IGptDecoderBatch9TensorPtrE" title="tensorrt_llm::runtime::IGptDecoderBatch::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_llm7runtime16IGptDecoderBatch14getCumLogProbsEv" title="Permalink 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_llm7runtime16IGptDecoderBatch12getParentIdsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime16IGptDecoderBatch12getParentIdsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime16IGptDecoderBatch12getParentIdsEv"></span><span id="tensorrt_llm::runtime::IGptDecoderBatch::getParentIdsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatch_1a1a37b811b0cc2a98507b57109844a06c"></span><span class="k"><span class="pre">virtual</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime16IGptDecoderBatch9TensorPtrE" title="tensorrt_llm::runtime::IGptDecoderBatch::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_llm7runtime16IGptDecoderBatch12getParentIdsEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime16IGptDecoderBatch10getNbStepsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime16IGptDecoderBatch10getNbStepsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime16IGptDecoderBatch10getNbStepsEv"></span><span id="tensorrt_llm::runtime::IGptDecoderBatch::getNbStepsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatch_1a935e778ac2c14276d940e09b9411e8ed"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime16IGptDecoderBatch10getNbStepsEv" title="Permalink 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_llm7runtime16IGptDecoderBatch16IGptDecoderBatchEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime16IGptDecoderBatch16IGptDecoderBatchEv"></span><span id="_CPPv2N12tensorrt_llm7runtime16IGptDecoderBatch16IGptDecoderBatchEv"></span><span id="tensorrt_llm::runtime::IGptDecoderBatch::IGptDecoderBatch"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IGptDecoderBatch_1abfc9e04a4f042f3c0220c84cff407ff3"></span><span class="sig-name descname"><span class="n"><span class="pre">IGptDecoderBatch</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_llm7runtime16IGptDecoderBatch16IGptDecoderBatchEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
<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="Permalink to this definition"></a><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="iGptDecoderBatch_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="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<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><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="#_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_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_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><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5InputE" title="Permalink 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_batch5Input4BaseE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch5Input4BaseE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch5Input4BaseE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Input_1afd98890b16c1fb49efd955e2f8cc74b5"></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="#_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><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input4BaseE" title="Permalink 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_batch5Input5InputE9TensorPtr">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch5Input5InputE9TensorPtr"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch5Input5InputE9TensorPtr"></span><span id="tensorrt_llm::runtime::decoder_batch::Input::Input__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Input_1a818425caddd990a62547b8cd80155f3b"></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">TensorPtr</span></span><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_llm7runtime13decoder_batch5Input5InputE9TensorPtr" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch5Input5InputE9TensorPtrRKNSt6vectorIbEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch5Input5InputE9TensorPtrRKNSt6vectorIbEE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch5Input5InputE9TensorPtrRKNSt6vectorIbEE"></span><span id="tensorrt_llm::runtime::decoder_batch::Input::Input__TensorPtr.std::vector:b:CR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Input_1a036ae2c032d83f9fa028bf7b1e56f72e"></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">TensorPtr</span></span><span class="w"> </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_batch5Input5InputE9TensorPtrRKNSt6vectorIbEE" title="Permalink 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_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="Permalink 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_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="Permalink 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_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_1a016b305f614d064ef624b5798d53dfc7"></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="w"> </span><span class="k"><span class="pre">const</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request9TensorPtrE" title="Permalink 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_batch7Request7RequestE9TensorPtrNSt8optionalI8SizeTypeEENSt8optionalI8SizeTypeEENSt8optionalI8SizeTypeEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch7Request7RequestE9TensorPtrNSt8optionalI8SizeTypeEENSt8optionalI8SizeTypeEENSt8optionalI8SizeTypeEE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch7Request7RequestE9TensorPtrNSt8optionalI8SizeTypeEENSt8optionalI8SizeTypeEENSt8optionalI8SizeTypeEE"></span><span id="tensorrt_llm::runtime::decoder_batch::Request::Request__TensorPtr.std::optional:SizeType:.std::optional:SizeType:.std::optional:SizeType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Request_1adc776675e0d99442e16cba4c39e7f7ab"></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_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="n sig-param"><span class="pre">ids</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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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="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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="p"><span class="pre">></span></span><span class="w"> </span><span class="n sig-param"><span class="pre">padId</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_batch7Request7RequestE9TensorPtrNSt8optionalI8SizeTypeEENSt8optionalI8SizeTypeEENSt8optionalI8SizeTypeEE" title="Permalink 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__TensorPtr"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Request_1a228f26f47eae229c4664895d797b77d1"></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">ids</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch7Request3idsE" title="Permalink 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:SizeType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Request_1a5ab1bf54050a431e7c26df7451f2f0f1"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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="Permalink 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:SizeType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Request_1ac1477f03a617c556bb264121200fcc0e"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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="Permalink 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="Permalink 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="Permalink 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="Permalink 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_batch5TokenE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch5TokenE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch5TokenE"></span><span id="tensorrt_llm::runtime::decoder_batch::Token"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Token"></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">Token</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime13decoder_batch5TokenE" title="Permalink 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_batch5Token5TokenERR9CudaEventRKNSt6vectorIbEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch5Token5TokenERR9CudaEventRKNSt6vectorIbEE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch5Token5TokenERR9CudaEventRKNSt6vectorIbEE"></span><span id="tensorrt_llm::runtime::decoder_batch::Token::Token__CudaEventRR.std::vector:b:CR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Token_1ab35d80ddc444d443a35ae642d3e9f862"></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">Token</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_batch5Token5TokenERR9CudaEventRKNSt6vectorIbEE" title="Permalink 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_batch5Token5eventE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch5Token5eventE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch5Token5eventE"></span><span id="tensorrt_llm::runtime::decoder_batch::Token::event__CudaEvent"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Token_1a8a4fa5d069ee58377e79ddde0955dbde"></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_batch5Token5eventE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime13decoder_batch5Token6activeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13decoder_batch5Token6activeE"></span><span id="_CPPv2N12tensorrt_llm7runtime13decoder_batch5Token6activeE"></span><span id="tensorrt_llm::runtime::decoder_batch::Token::active__std::vector:b:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Token_1a92563b2a16378f689f9d600d8712f8fd"></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_batch5Token6activeE" title="Permalink 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="Permalink 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_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="Permalink 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_1IGptDecoderBatch"><span class="std std-ref">tensorrt_llm::runtime::IGptDecoderBatch</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="Permalink 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="Permalink 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_llm7runtime19IStatefulGptDecoder5setupE8SizeType8SizeType8SizeTypeN8nvinfer18DataTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime19IStatefulGptDecoder5setupE8SizeType8SizeType8SizeTypeN8nvinfer18DataTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime19IStatefulGptDecoder5setupE8SizeType8SizeType8SizeTypeN8nvinfer18DataTypeE"></span><span id="tensorrt_llm::runtime::IStatefulGptDecoder::setup__SizeType.SizeType.SizeType.nvinfer1::DataType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IStatefulGptDecoder_1a40b70b01f26c50f974a46163300c6d9a"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxBatchSize</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">maxBeamWidth</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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">dtype</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_llm7runtime19IStatefulGptDecoder5setupE8SizeType8SizeType8SizeTypeN8nvinfer18DataTypeE" title="Permalink 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_1a65d6bb6723258d57f5d58483749c2004"><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_llm7runtime19IStatefulGptDecoder8newBatchERK15GenerationInputRK14SamplingConfig">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime19IStatefulGptDecoder8newBatchERK15GenerationInputRK14SamplingConfig"></span><span id="_CPPv2N12tensorrt_llm7runtime19IStatefulGptDecoder8newBatchERK15GenerationInputRK14SamplingConfig"></span><span id="tensorrt_llm::runtime::IStatefulGptDecoder::newBatch__GenerationInputCR.SamplingConfigCR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IStatefulGptDecoder_1a6dfda7e2cc24d12a387dbef8801b2211"></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_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="p"><span class="pre">=</span></span><span class="w"> </span><span class="m"><span class="pre">0</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime19IStatefulGptDecoder8newBatchERK15GenerationInputRK14SamplingConfig" title="Permalink 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="Permalink 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_llm7runtime19IStatefulGptDecoder14isFinishedSyncEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime19IStatefulGptDecoder14isFinishedSyncEv"></span><span id="_CPPv2N12tensorrt_llm7runtime19IStatefulGptDecoder14isFinishedSyncEv"></span><span id="tensorrt_llm::runtime::IStatefulGptDecoder::isFinishedSync"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IStatefulGptDecoder_1a4840aa6124a2590af2b575faf84bfbf7"></span><span class="k"><span class="pre">virtual</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">isFinishedSync</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_llm7runtime19IStatefulGptDecoder14isFinishedSyncEv" title="Permalink 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 and return whether all sequences have finished. </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_1a65d6bb6723258d57f5d58483749c2004"></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">bool</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="Permalink 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_llm7runtime19IStatefulGptDecoder17getFinalOutputIdsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime19IStatefulGptDecoder17getFinalOutputIdsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime19IStatefulGptDecoder17getFinalOutputIdsEv"></span><span id="tensorrt_llm::runtime::IStatefulGptDecoder::getFinalOutputIdsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IStatefulGptDecoder_1a54e4feeae207c8426d04011e3ca58ef7"></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">getFinalOutputIds</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_llm7runtime19IStatefulGptDecoder17getFinalOutputIdsEv" title="Permalink to this definition"></a><br /></dt>
|
||
<dd><p>Gather final results for all requests. </p>
|
||
</dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime19IStatefulGptDecoder12getOutputIdsEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime19IStatefulGptDecoder12getOutputIdsEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime19IStatefulGptDecoder12getOutputIdsEv"></span><span id="tensorrt_llm::runtime::IStatefulGptDecoder::getOutputIdsC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IStatefulGptDecoder_1a3f98067a78b64a319acb4a8b667e6249"></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">getOutputIds</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_llm7runtime19IStatefulGptDecoder12getOutputIdsEv" title="Permalink 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_llm7runtime19IStatefulGptDecoder12getNewTokensEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime19IStatefulGptDecoder12getNewTokensEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime19IStatefulGptDecoder12getNewTokensEv"></span><span id="tensorrt_llm::runtime::IStatefulGptDecoder::getNewTokensC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IStatefulGptDecoder_1a2806684ae6ae3a9f4d96042c881d9993"></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><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_llm7runtime19IStatefulGptDecoder12getNewTokensEv" title="Permalink 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], latests generated tokens (per beam), 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="Permalink 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>
|
||
|
||
</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="Permalink 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="Permalink 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="Permalink to this definition"></a><br /></dt>
|
||
<dd><p>Subclassed by <a class="reference internal" href="#classtensorrt__llm_1_1runtime_1_1decoder__batch_1_1Input"><span class="std std-ref">tensorrt_llm::runtime::decoder_batch::Input</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_llm7runtime7decoder5Input9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7decoder5Input9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime7decoder5Input9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1decoder_1_1Input_1a00dc3503cac2813698c97486fba9ca51"></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="w"> </span><span class="k"><span class="pre">const</span></span><span class="p"><span class="pre">></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime7decoder5Input9TensorPtrE" title="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="itensor-h">
|
||
<h2>iTensor.h<a class="headerlink" href="#itensor-h" title="Permalink 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="Permalink 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_1ab843bef84b07b782c05c682399ee2b1a"></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="Permalink 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_1a2a38821feb0276616dde2d32020ef451"></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="Permalink to this definition"></a><br /></dt>
|
||
<dd><p>Utility function to print a tensor with its shape. </p>
|
||
</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="#_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_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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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_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_1a9f2d4a0903e581784f5ed94d8ed1d58d"></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="Permalink 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="_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="Permalink 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_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="Permalink to this definition"></a><br /></dt>
|
||
<dd></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="Permalink 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_1a943f0acaa8c746fa85980f4e13512a2c"></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="Permalink 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_llm7runtime7ITensor7squeezeE8SizeType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor7squeezeE8SizeType"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor7squeezeE8SizeType"></span><span id="tensorrt_llm::runtime::ITensor::squeeze__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a5ca07f1cc675136634c6a435b5b8f8f3"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime7ITensor7squeezeE8SizeType" title="Permalink to this definition"></a><br /></dt>
|
||
<dd><p>Removes the given <em>unit</em> dimensions from this tensor. </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_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="Permalink 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="Permalink 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_llm7runtime7ITensor7squeezeERK5Shape8SizeType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor7squeezeERK5Shape8SizeType"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor7squeezeERK5Shape8SizeType"></span><span id="tensorrt_llm::runtime::ITensor::squeeze__ShapeCR.SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a7c9ddd4f8b1eaa7b1a6276c49b2c8672"></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">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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime7ITensor7squeezeERK5Shape8SizeType" title="Permalink 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_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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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_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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime7ITensor9makeShapeERKNSt16initializer_listI8SizeTypeEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime7ITensor9makeShapeERKNSt16initializer_listI8SizeTypeEE"></span><span id="_CPPv2N12tensorrt_llm7runtime7ITensor9makeShapeERKNSt16initializer_listI8SizeTypeEE"></span><span id="tensorrt_llm::runtime::ITensor::makeShape__std::initializer_list:SizeType:CR"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1ITensor_1a7ba94f17049ee1d6048ae69be48ecb67"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_listI8SizeTypeEE" title="Permalink 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="Permalink 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>
|
||
|
||
</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="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="ipcutils-h">
|
||
<h2>ipcUtils.h<a class="headerlink" href="#ipcutils-h" title="Permalink 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_llm7runtime13setPeerAccessE11WorldConfigb">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime13setPeerAccessE11WorldConfigb"></span><span id="_CPPv2N12tensorrt_llm7runtime13setPeerAccessE11WorldConfigb"></span><span id="tensorrt_llm::runtime::setPeerAccess__WorldConfig.b"></span><span class="target" id="ipcUtils_8h_1a16050a9ed921a12d13a273478086f95c"></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">setPeerAccess</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="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">enable</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_llm7runtime13setPeerAccessE11WorldConfigb" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
<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="Permalink 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_llm7runtime9IpcMemory9TensorPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9IpcMemory9TensorPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime9IpcMemory9TensorPtrE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IpcMemory_1a1c17fa36835ec34cb3cef927e85967eb"></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_llm7runtime9IpcMemory9TensorPtrE" title="Permalink 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_llm7runtime9IpcMemory9IpcMemoryE11WorldConfigNSt6size_tE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9IpcMemory9IpcMemoryE11WorldConfigNSt6size_tE"></span><span id="_CPPv2N12tensorrt_llm7runtime9IpcMemory9IpcMemoryE11WorldConfigNSt6size_tE"></span><span id="tensorrt_llm::runtime::IpcMemory::IpcMemory__WorldConfig.std::s"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IpcMemory_1ab60128a164106fe08d754357214b5b82"></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_llm7runtime11WorldConfigE" title="tensorrt_llm::runtime::WorldConfig"><span class="n"><span class="pre">WorldConfig</span></span></a><span class="w"> </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">size_t</span></span><span class="w"> </span><span class="n sig-param"><span class="pre">bufferSize</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory9IpcMemoryE11WorldConfigNSt6size_tE" title="Permalink 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="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4NK12tensorrt_llm7runtime9IpcMemory17getCommPtrsTensorEv">
|
||
<span id="_CPPv3NK12tensorrt_llm7runtime9IpcMemory17getCommPtrsTensorEv"></span><span id="_CPPv2NK12tensorrt_llm7runtime9IpcMemory17getCommPtrsTensorEv"></span><span id="tensorrt_llm::runtime::IpcMemory::getCommPtrsTensorC"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IpcMemory_1a8bcbdce69d5e3de95a0f742e430f822c"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><span class="k"><span class="pre">const</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="p"><span class="pre">&</span></span><span class="sig-name descname"><span class="n"><span class="pre">getCommPtrsTensor</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_llm7runtime9IpcMemory17getCommPtrsTensorEv" title="Permalink 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_1a4146075f3f6beba4eab4a765506be9bd"></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">size_t</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="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="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="Permalink 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_llm7runtime9IpcMemory17allocateIpcMemoryEv">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9IpcMemory17allocateIpcMemoryEv"></span><span id="_CPPv2N12tensorrt_llm7runtime9IpcMemory17allocateIpcMemoryEv"></span><span id="tensorrt_llm::runtime::IpcMemory::allocateIpcMemory"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IpcMemory_1a68e4ab83a9194bdd5cf59a1d4ee68a80"></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="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory17allocateIpcMemoryEv" title="Permalink 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="Permalink 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_llm7runtime9IpcMemory12mWorldConfigE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9IpcMemory12mWorldConfigE"></span><span id="_CPPv2N12tensorrt_llm7runtime9IpcMemory12mWorldConfigE"></span><span id="tensorrt_llm::runtime::IpcMemory::mWorldConfig__WorldConfig"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IpcMemory_1aab66f175e30581b311c654816aa85ea4"></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_llm7runtime9IpcMemory12mWorldConfigE" title="Permalink 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="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9IpcMemory11mBufferSizeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9IpcMemory11mBufferSizeE"></span><span id="_CPPv2N12tensorrt_llm7runtime9IpcMemory11mBufferSizeE"></span><span id="tensorrt_llm::runtime::IpcMemory::mBufferSize__std::s"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IpcMemory_1a95b1167806fe77f1f22810ac7477d27d"></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">mBufferSize</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory11mBufferSizeE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp var">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime9IpcMemory10mBufferPtrE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime9IpcMemory10mBufferPtrE"></span><span id="_CPPv2N12tensorrt_llm7runtime9IpcMemory10mBufferPtrE"></span><span id="tensorrt_llm::runtime::IpcMemory::mBufferPtr__voidP"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1IpcMemory_1adf62ec2d27e1382c5414188fd3982f8b"></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">mBufferPtr</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime9IpcMemory10mBufferPtrE" title="Permalink 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="Permalink 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="Permalink 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_llm7runtime14MemoryCounters8SizeTypeE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14MemoryCounters8SizeTypeE"></span><span id="_CPPv2N12tensorrt_llm7runtime14MemoryCounters8SizeTypeE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a5aa02d808601c67ba934f8a0a9534d6e"></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">SizeType</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_llm7runtime14MemoryCounters8SizeTypeE" title="Permalink 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="Permalink 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="Permalink 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_1a09119207f2dfca61e4e7bd5a66c36654"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters8SizeTypeE" title="tensorrt_llm::runtime::MemoryCounters::SizeType"><span class="n"><span class="pre">SizeType</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="Permalink 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_1af643b8967c9f99d8efd647d7b3efe281"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters8SizeTypeE" title="tensorrt_llm::runtime::MemoryCounters::SizeType"><span class="n"><span class="pre">SizeType</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="Permalink 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_1ae37c0265530ae4fac20d31da5712d3c0"></span><span class="k"><span class="pre">inline</span></span><span class="w"> </span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters8SizeTypeE" title="tensorrt_llm::runtime::MemoryCounters::SizeType"><span class="n"><span class="pre">SizeType</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="Permalink 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="Permalink 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="Permalink 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="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I_10MemoryTypeEN12tensorrt_llm7runtime14MemoryCounters8allocateEv8SizeType">
|
||
<span id="_CPPv3I_10MemoryTypeEN12tensorrt_llm7runtime14MemoryCounters8allocateE8SizeType"></span><span id="_CPPv2I_10MemoryTypeEN12tensorrt_llm7runtime14MemoryCounters8allocateE8SizeType"></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_1a891aab52104e74b95138f4bf20c111c2"></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_llm7runtime14MemoryCounters8SizeTypeE" title="tensorrt_llm::runtime::MemoryCounters::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime14MemoryCounters8allocateEv8SizeType" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14MemoryCounters8allocateE10MemoryType8SizeType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14MemoryCounters8allocateE10MemoryType8SizeType"></span><span id="_CPPv2N12tensorrt_llm7runtime14MemoryCounters8allocateE10MemoryType8SizeType"></span><span id="tensorrt_llm::runtime::MemoryCounters::allocate__MemoryType.SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a0b437c35c4af66d2f2b2379a40c136d6"></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_llm7runtime14MemoryCounters8SizeTypeE" title="tensorrt_llm::runtime::MemoryCounters::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime14MemoryCounters8allocateE10MemoryType8SizeType" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4I_10MemoryTypeEN12tensorrt_llm7runtime14MemoryCounters10deallocateEv8SizeType">
|
||
<span id="_CPPv3I_10MemoryTypeEN12tensorrt_llm7runtime14MemoryCounters10deallocateE8SizeType"></span><span id="_CPPv2I_10MemoryTypeEN12tensorrt_llm7runtime14MemoryCounters10deallocateE8SizeType"></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_1a6301480be8d33d2a3e7dbd54119467c9"></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_llm7runtime14MemoryCounters8SizeTypeE" title="tensorrt_llm::runtime::MemoryCounters::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime14MemoryCounters10deallocateEv8SizeType" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14MemoryCounters10deallocateE10MemoryType8SizeType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14MemoryCounters10deallocateE10MemoryType8SizeType"></span><span id="_CPPv2N12tensorrt_llm7runtime14MemoryCounters10deallocateE10MemoryType8SizeType"></span><span id="tensorrt_llm::runtime::MemoryCounters::deallocate__MemoryType.SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a4cbe26fb0d89352c1de2cb6f63220aa2"></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_llm7runtime14MemoryCounters8SizeTypeE" title="tensorrt_llm::runtime::MemoryCounters::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime14MemoryCounters10deallocateE10MemoryType8SizeType" title="Permalink 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_1a5fd849b9f6dd583910a3aa4acb8ea391"></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_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="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime14MemoryCounters13bytesToStringE8SizeTypei">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14MemoryCounters13bytesToStringE8SizeTypei"></span><span id="_CPPv2N12tensorrt_llm7runtime14MemoryCounters13bytesToStringE8SizeTypei"></span><span id="tensorrt_llm::runtime::MemoryCounters::bytesToString__SizeType.i"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1ad3e1b845f41a4245556557b1393881f9"></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_llm7runtime14MemoryCounters8SizeTypeE" title="tensorrt_llm::runtime::MemoryCounters::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime14MemoryCounters13bytesToStringE8SizeTypei" title="Permalink 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="Permalink 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__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a641ecc1e58aa222f4a7e269aa589e70f"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters8SizeTypeE" title="tensorrt_llm::runtime::MemoryCounters::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><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="Permalink 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__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a27771c9783292c51de1b8a59aca33d71"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters8SizeTypeE" title="tensorrt_llm::runtime::MemoryCounters::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><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="Permalink 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__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a7eaaa821b0f7cbf8f1a2264e5f1b1c25"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters8SizeTypeE" title="tensorrt_llm::runtime::MemoryCounters::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><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="Permalink 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__DiffType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a091075678e2a77da992d619ccc27ee69"></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">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="Permalink 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__DiffType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a55b97586df47d83ece390da5b9da7ace"></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">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="Permalink 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__DiffType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a3ac02c1fc74dd309e7680be2c27ecbf0"></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">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="Permalink 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_llm7runtime14MemoryCounters9mInstanceE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14MemoryCounters9mInstanceE"></span><span id="_CPPv2N12tensorrt_llm7runtime14MemoryCounters9mInstanceE"></span><span id="tensorrt_llm::runtime::MemoryCounters::mInstance__MemoryCounters"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1MemoryCounters_1a6de4abe7564f08c88fedabcff29ffeca"></span><span class="k"><span class="pre">static</span></span><span class="w"> </span><span class="k"><span class="pre">thread_local</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="sig-name descname"><span class="n"><span class="pre">mInstance</span></span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime14MemoryCounters9mInstanceE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
</div>
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</dd></dl>
|
||
|
||
</section>
|
||
<section id="samplingconfig-h">
|
||
<h2>samplingConfig.h<a class="headerlink" href="#samplingconfig-h" title="Permalink 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_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="Permalink 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_llm7runtime14SamplingConfig14SamplingConfigE8SizeType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime14SamplingConfig14SamplingConfigE8SizeType"></span><span id="_CPPv2N12tensorrt_llm7runtime14SamplingConfig14SamplingConfigE8SizeType"></span><span id="tensorrt_llm::runtime::SamplingConfig::SamplingConfig__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1a1b84a281c4413d5a4df922e9074d85ac"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime14SamplingConfig14SamplingConfigE8SizeType" title="Permalink 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__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1a871887c0857fa88f45fadcd659c20824"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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="Permalink 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="Permalink 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:SizeType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1aacc380bc9c6ff46fd2edd112b1eac9c0"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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="Permalink 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="Permalink 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="Permalink 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:SizeType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1ab1588865e1c6e97bfebfe12b7ee34e37"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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="Permalink 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="Permalink 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:unsigned-l-l:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1a6468682988b1992a5be330aaadc9641e"></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">unsigned</span></span><span class="w"> </span><span class="kt"><span class="pre">long</span></span><span class="w"> </span><span class="kt"><span class="pre">long</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="Permalink 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="Permalink 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="Permalink 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:SizeType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1ae2a3d686956933a6938925309496a33e"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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="Permalink 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__std::optional:FloatType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1aa97844dc62b1cf9eac23d606697a575c"></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_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="Permalink 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__std::optional:FloatType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1SamplingConfig_1a902a829073ae4bc0745374e3f5d07a13"></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_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="Permalink 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="Permalink 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="Permalink 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="Permalink 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><span class="n"><span class="pre">ILogger</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime10TllmLoggerE" title="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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="Permalink 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_llm7runtime11WorldConfig11WorldConfigE8SizeType8SizeType8SizeType8SizeType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11WorldConfig11WorldConfigE8SizeType8SizeType8SizeType8SizeType"></span><span id="_CPPv2N12tensorrt_llm7runtime11WorldConfig11WorldConfigE8SizeType8SizeType8SizeType8SizeType"></span><span id="tensorrt_llm::runtime::WorldConfig::WorldConfig__SizeType.SizeType.SizeType.SizeTypeCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1a24619a7cd8eab9da4347680e248b922a"></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">WorldConfig</span></span></span><span class="sig-paren">(</span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfig11WorldConfigE8SizeType8SizeType8SizeType8SizeType" title="Permalink 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_1ab015227c42d7122c0e33f22e03ac8228"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><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="Permalink 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_1a389ba0bc53cd7fe0259bdb2124ee156f"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><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="Permalink 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_1af3627cbcb19d1519338fea2e43335633"></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">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="Permalink 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_1ab90953f5e50d6fdef2d3c76ad8295f98"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><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="Permalink 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_1a37ec8744e7dc7ff6211c319a33ecc628"></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">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="Permalink 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_1ab99195015c992ba7a735c7880cc62bb6"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><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="Permalink 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_1a32f320679d97d1a278e84bb21af89494"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><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="Permalink 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::getDeviceCCE"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1a35f8f78f02108160d180c0b70f78b414"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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="Permalink 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_1a619d3cbd60a9cad3d0232844e8693739"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><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="Permalink 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_1ab00da06440a9b4f3305d9811dcd7b61b"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><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="Permalink 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_1a6a5a2512cce31560ae828d7fd1968c22"></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">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="Permalink 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_1a7e9cc05c32f9a1edf82d4017e34a3bef"></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">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="Permalink 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_1a79368880ccca35154c16cbf409c97430"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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="Permalink 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_llm7runtime11WorldConfig11validConfigERN8nvinfer17ILoggerE8SizeType8SizeType">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11WorldConfig11validConfigERN8nvinfer17ILoggerE8SizeType8SizeType"></span><span id="_CPPv2N12tensorrt_llm7runtime11WorldConfig11validConfigERN8nvinfer17ILoggerE8SizeType8SizeType"></span><span id="tensorrt_llm::runtime::WorldConfig::validConfig__nvinfer1::ILoggerR.SizeType.SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1ab34b3cafb821617e539a8d736c09e3f5"></span><span class="k"><span class="pre">static</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">validConfig</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">ILogger</span></span><span class="w"> </span><span class="p"><span class="pre">&</span></span><span class="n sig-param"><span class="pre">logger</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">tensorParallelism</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><span class="w"> </span><span class="n sig-param"><span class="pre">pipelineParallelism</span></span><span class="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfig11validConfigERN8nvinfer17ILoggerE8SizeType8SizeType" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11WorldConfig3mpiERN8nvinfer17ILoggerE8SizeTypeNSt8optionalI8SizeTypeEENSt8optionalI8SizeTypeEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11WorldConfig3mpiERN8nvinfer17ILoggerE8SizeTypeNSt8optionalI8SizeTypeEENSt8optionalI8SizeTypeEE"></span><span id="_CPPv2N12tensorrt_llm7runtime11WorldConfig3mpiERN8nvinfer17ILoggerE8SizeTypeNSt8optionalI8SizeTypeEENSt8optionalI8SizeTypeEE"></span><span id="tensorrt_llm::runtime::WorldConfig::mpi__nvinfer1::ILoggerR.SizeType.std::optional:SizeType:.std::optional:SizeType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1ad8ec600840b45582dccbc01eb7926ff0"></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="#_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="n sig-param"><span class="pre">logger</span></span>, <a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfig3mpiERN8nvinfer17ILoggerE8SizeTypeNSt8optionalI8SizeTypeEENSt8optionalI8SizeTypeEE" title="Permalink to this definition"></a><br /></dt>
|
||
<dd></dd></dl>
|
||
|
||
<dl class="cpp function">
|
||
<dt class="sig sig-object cpp" id="_CPPv4N12tensorrt_llm7runtime11WorldConfig3mpiE8SizeTypeNSt8optionalI8SizeTypeEENSt8optionalI8SizeTypeEE">
|
||
<span id="_CPPv3N12tensorrt_llm7runtime11WorldConfig3mpiE8SizeTypeNSt8optionalI8SizeTypeEENSt8optionalI8SizeTypeEE"></span><span id="_CPPv2N12tensorrt_llm7runtime11WorldConfig3mpiE8SizeTypeNSt8optionalI8SizeTypeEENSt8optionalI8SizeTypeEE"></span><span id="tensorrt_llm::runtime::WorldConfig::mpi__SizeType.std::optional:SizeType:.std::optional:SizeType:"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1ab0651f2698fe41426584b3ab69529460"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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="sig-paren">)</span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfig3mpiE8SizeTypeNSt8optionalI8SizeTypeEENSt8optionalI8SizeTypeEE" title="Permalink 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__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1a2d68f68654533c7b0b9be92dae183082"></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_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</span></span></a><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">8</span></span><a class="headerlink" href="#_CPPv4N12tensorrt_llm7runtime11WorldConfig19kDefaultGpusPerNodeE" title="Permalink 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__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1a60d3f95841a56216582ff4d1749841db"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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="Permalink 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__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1a455c07380f4ab9f5d299de40011d03be"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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="Permalink 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__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1a834b8b05fadab2839ad94c0389385330"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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="Permalink 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__SizeType"></span><span class="target" id="classtensorrt__llm_1_1runtime_1_1WorldConfig_1a3c6d093540443a20a3204d0129d8be79"></span><a class="reference internal" href="#_CPPv4N12tensorrt_llm7runtime8SizeTypeE" title="tensorrt_llm::runtime::SizeType"><span class="n"><span class="pre">SizeType</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="Permalink 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="../python-api/tensorrt_llm.runtime.html" class="btn btn-neutral float-left" title="Runtime" accesskey="p" rel="prev"><span class="fa fa-arrow-circle-left" aria-hidden="true"></span> Previous</a>
|
||
</div>
|
||
|
||
<hr/>
|
||
|
||
<div role="contentinfo">
|
||
<p>© Copyright 2023, NVidia.</p>
|
||
</div>
|
||
|
||
Built with <a href="https://www.sphinx-doc.org/">Sphinx</a> using a
|
||
<a href="https://github.com/readthedocs/sphinx_rtd_theme">theme</a>
|
||
provided by <a href="https://readthedocs.org">Read the Docs</a>.
|
||
|
||
|
||
</footer>
|
||
</div>
|
||
</div>
|
||
</section>
|
||
</div>
|
||
<script>
|
||
jQuery(function () {
|
||
SphinxRtdTheme.Navigation.enable(true);
|
||
});
|
||
</script>
|
||
|
||
</body>
|
||
</html> |