kanshan/TensorRT-LLMs
1
0
Fork 0
mirror of https://github.com/NVIDIA/TensorRT-LLM.git synced 2026-01-14 06:27:45 +08:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity
42d0585493
TensorRT-LLMs/architecture/core-concepts.html
nv-guomingz 42d0585493
doc:fix llm api reference blank issue. (#2669)
2025-01-08 14:45:46 +08:00

582 lines
55 KiB
HTML
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

<!DOCTYPE html>
<html class="writer-html5" lang="en" data-content_root="../">
<head>
<meta charset="utf-8" /><meta name="viewport" content="width=device-width, initial-scale=1" />
<meta name="viewport" content="width=device-width, initial-scale=1.0" />
<title>Model Definition &mdash; tensorrt_llm documentation</title>
<link rel="stylesheet" type="text/css" href="../_static/pygments.css?v=80d5e7a1" />
<link rel="stylesheet" type="text/css" href="../_static/css/theme.css?v=e59714d7" />
<link rel="stylesheet" type="text/css" href="../_static/copybutton.css?v=76b2166b" />
<script src="../_static/jquery.js?v=5d32c60e"></script>
<script src="../_static/_sphinx_javascript_frameworks_compat.js?v=2cd50e6c"></script>
<script src="../_static/documentation_options.js?v=5929fcd5"></script>
<script src="../_static/doctools.js?v=9bcbadda"></script>
<script src="../_static/sphinx_highlight.js?v=dc90522c"></script>
<script src="../_static/clipboard.min.js?v=a7894cd8"></script>
<script src="../_static/copybutton.js?v=65e89d2a"></script>
<script src="../_static/js/theme.js"></script>
<link rel="index" title="Index" href="../genindex.html" />
<link rel="search" title="Search" href="../search.html" />
<link rel="next" title="TensorRT-LLM Checkpoint" href="checkpoint.html" />
<link rel="prev" title="TensorRT-LLM Architecture" href="overview.html" />
</head>
<body class="wy-body-for-nav">
<div class="wy-grid-for-nav">
<nav data-toggle="wy-nav-shift" class="wy-nav-side">
<div class="wy-side-scroll">
<div class="wy-side-nav-search" >
<a href="../index.html" class="icon icon-home">
tensorrt_llm
</a>
<div role="search">
<form id="rtd-search-form" class="wy-form" action="../search.html" method="get">
<input type="text" name="q" placeholder="Search docs" aria-label="Search docs" />
<input type="hidden" name="check_keywords" value="yes" />
<input type="hidden" name="area" value="default" />
</form>
</div>
</div><div class="wy-menu wy-menu-vertical" data-spy="affix" role="navigation" aria-label="Navigation menu">
<p class="caption" role="heading"><span class="caption-text">Getting Started</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../overview.html">Overview</a></li>
<li class="toctree-l1"><a class="reference internal" href="../quick-start-guide.html">Quick Start Guide</a></li>
<li class="toctree-l1"><a class="reference internal" href="../key-features.html">Key Features</a></li>
<li class="toctree-l1"><a class="reference internal" href="../release-notes.html">Release Notes</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Installation</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../installation/linux.html">Installing on Linux</a></li>
<li class="toctree-l1"><a class="reference internal" href="../installation/build-from-source-linux.html">Building from Source Code on Linux</a></li>
<li class="toctree-l1"><a class="reference internal" href="../installation/windows.html">Installing on Windows</a></li>
<li class="toctree-l1"><a class="reference internal" href="../installation/build-from-source-windows.html">Building from Source Code on Windows</a></li>
<li class="toctree-l1"><a class="reference internal" href="../installation/grace-hopper.html">Installing on Grace Hopper</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">LLM API</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../llm-api/index.html">API Introduction</a></li>
<li class="toctree-l1"><a class="reference internal" href="../llm-api/reference.html">API Reference</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">LLM API Examples</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../llm-api-examples/index.html">LLM Examples Introduction</a></li>
<li class="toctree-l1"><a class="reference internal" href="../llm-api-examples/customization.html">Common Customizations</a></li>
<li class="toctree-l1"><a class="reference internal" href="../llm-api-examples/llm_api_examples.html">Examples</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Model Definition API</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../python-api/tensorrt_llm.layers.html">Layers</a></li>
<li class="toctree-l1"><a class="reference internal" href="../python-api/tensorrt_llm.functional.html">Functionals</a></li>
<li class="toctree-l1"><a class="reference internal" href="../python-api/tensorrt_llm.models.html">Models</a></li>
<li class="toctree-l1"><a class="reference internal" href="../python-api/tensorrt_llm.plugin.html">Plugin</a></li>
<li class="toctree-l1"><a class="reference internal" href="../python-api/tensorrt_llm.quantization.html">Quantization</a></li>
<li class="toctree-l1"><a class="reference internal" href="../python-api/tensorrt_llm.runtime.html">Runtime</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">C++ API</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../_cpp_gen/executor.html">Executor</a></li>
<li class="toctree-l1"><a class="reference internal" href="../_cpp_gen/runtime.html">Runtime</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Command-Line Reference</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../commands/trtllm-build.html">trtllm-build</a></li>
<li class="toctree-l1"><a class="reference internal" href="../commands/trtllm-serve.html">trtllm-serve</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Architecture</span></p>
<ul class="current">
<li class="toctree-l1"><a class="reference internal" href="overview.html">TensorRT-LLM Architecture</a></li>
<li class="toctree-l1 current"><a class="current reference internal" href="#">Model Definition</a></li>
<li class="toctree-l1"><a class="reference internal" href="#compilation">Compilation</a><ul>
<li class="toctree-l2"><a class="reference internal" href="#tensorrt-compiler">TensorRT Compiler</a></li>
<li class="toctree-l2"><a class="reference internal" href="#model-engine">Model Engine</a></li>
<li class="toctree-l2"><a class="reference internal" href="#weight-bindings">Weight Bindings</a></li>
<li class="toctree-l2"><a class="reference internal" href="#pattern-matching-and-fusion">Pattern-Matching and Fusion</a></li>
<li class="toctree-l2"><a class="reference internal" href="#plugins">Plugins</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="#runtime">Runtime</a></li>
<li class="toctree-l1"><a class="reference internal" href="#multi-gpu-and-multi-node-support">Multi-GPU and Multi-Node Support</a><ul>
<li class="toctree-l2"><a class="reference internal" href="#examples">Examples</a><ul>
<li class="toctree-l3"><a class="reference internal" href="#llama-3-1-70b">Llama 3.1 70B</a></li>
<li class="toctree-l3"><a class="reference internal" href="#llama-3-1-405b">Llama 3.1 405B</a></li>
</ul>
</li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="checkpoint.html">TensorRT-LLM Checkpoint</a></li>
<li class="toctree-l1"><a class="reference internal" href="workflow.html">TensorRT-LLM Build Workflow</a></li>
<li class="toctree-l1"><a class="reference internal" href="add-model.html">Adding a Model</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Advanced</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../advanced/gpt-attention.html">Multi-Head, Multi-Query, and Group-Query Attention</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/gpt-runtime.html">C++ GPT Runtime</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/executor.html">Executor API</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/graph-rewriting.html">Graph Rewriting Module</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/inference-request.html">Inference Request</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/inference-request.html#responses">Responses</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/lora.html">Run gpt-2b + LoRA using GptManager / cpp runtime</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/expert-parallelism.html">Expert Parallelism in TensorRT-LLM</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/kv-cache-reuse.html">KV cache reuse</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/speculative-decoding.html">Speculative Sampling</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/disaggregated-service.html">Disaggregated-Service (experimental)</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Performance</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../performance/perf-overview.html">Overview</a></li>
<li class="toctree-l1"><a class="reference internal" href="../performance/perf-benchmarking.html">Benchmarking</a></li>
<li class="toctree-l1"><a class="reference internal" href="../performance/perf-best-practices.html">Best Practices</a></li>
<li class="toctree-l1"><a class="reference internal" href="../performance/perf-analysis.html">Performance Analysis</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Reference</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../reference/troubleshooting.html">Troubleshooting</a></li>
<li class="toctree-l1"><a class="reference internal" href="../reference/support-matrix.html">Support Matrix</a></li>
<li class="toctree-l1"><a class="reference internal" href="../reference/precision.html">Numerical Precision</a></li>
<li class="toctree-l1"><a class="reference internal" href="../reference/memory.html">Memory Usage of TensorRT-LLM</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Blogs</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../blogs/H100vsA100.html">H100 has 4.6x A100 Performance in TensorRT-LLM, achieving 10,000 tok/s at 100ms to first token</a></li>
<li class="toctree-l1"><a class="reference internal" href="../blogs/H200launch.html">H200 achieves nearly 12,000 tokens/sec on Llama2-13B with TensorRT-LLM</a></li>
<li class="toctree-l1"><a class="reference internal" href="../blogs/Falcon180B-H200.html">Falcon-180B on a single H200 GPU with INT4 AWQ, and 6.7x faster Llama-70B over A100</a></li>
<li class="toctree-l1"><a class="reference internal" href="../blogs/quantization-in-TRT-LLM.html">Speed up inference with SOTA quantization techniques in TRT-LLM</a></li>
<li class="toctree-l1"><a class="reference internal" href="../blogs/XQA-kernel.html">New XQA-kernel provides 2.4x more Llama-70B throughput within the same latency budget</a></li>
</ul>
</div>
</div>
</nav>
<section data-toggle="wy-nav-shift" class="wy-nav-content-wrap"><nav class="wy-nav-top" aria-label="Mobile navigation menu" >
<i data-toggle="wy-nav-top" class="fa fa-bars"></i>
<a href="../index.html">tensorrt_llm</a>
</nav>
<div class="wy-nav-content">
<div class="rst-content">
<div role="navigation" aria-label="Page navigation">
<ul class="wy-breadcrumbs">
<li><a href="../index.html" class="icon icon-home" aria-label="Home"></a></li>
<li class="breadcrumb-item active">Model Definition</li>
<li class="wy-breadcrumbs-aside">
<a href="../_sources/architecture/core-concepts.md.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="model-definition">
<span id="core-concepts"></span><h1>Model Definition<a class="headerlink" href="#model-definition" title="Link to this heading"></a></h1>
<p>TensorRT-LLM has a Model Definition API that can be used to define
Large Language Models. This API is built on top of the powerful
<a class="reference external" href="https://docs.nvidia.com/deeplearning/tensorrt/api/python_api/index.html#">TensorRT Python API</a>
to create graph representations of deep neural networks in TensorRT. To become
familiar with the core concepts of the TensorRT API, refer to the
<a class="reference external" href="https://docs.nvidia.com/deeplearning/tensorrt/api/python_api/coreConcepts.html">Core Concepts</a>
section of the TensorRT documentation before proceeding further.</p>
<p>In TensorRT-LLM, the <a class="reference external" href="https://github.com/NVIDIA/TensorRT-LLM/tree/rel/tensorrt_llm/builder.py"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm.Builder</span></code></a> class
contains a
<a class="reference external" href="https://docs.nvidia.com/deeplearning/tensorrt/api/python_api/infer/Core/Builder.html#tensorrt.Builder"><code class="docutils literal notranslate"><span class="pre">tensorrt.Builder</span></code></a>
object. That instance is used in the <code class="docutils literal notranslate"><span class="pre">tensorrt_llm.Builder.create_network</span></code>
method to create an instance of the
<a class="reference external" href="https://docs.nvidia.com/deeplearning/tensorrt/api/python_api/infer/Graph/Network.html#tensorrt.INetworkDefinition"><code class="docutils literal notranslate"><span class="pre">tensorrt.INetworkDefinition</span></code></a>
class. The <code class="docutils literal notranslate"><span class="pre">INetworkDefinition</span></code> object can then be populated using the free
functions defined in the
<a class="reference external" href="https://github.com/NVIDIA/TensorRT-LLM/tree/rel/tensorrt_llm/functional.py"><code class="docutils literal notranslate"><span class="pre">tensorrt_llm.functional</span></code></a>.</p>
<p>A simple example of such a free function is <code class="docutils literal notranslate"><span class="pre">tensorrt_llm.activation</span></code> that inserts a
<a class="reference external" href="https://docs.nvidia.com/deeplearning/tensorrt/api/python_api/infer/Graph/Layers.html#tensorrt.IActivationLayer"><code class="docutils literal notranslate"><span class="pre">tensorrt.IActivationLayer</span></code></a>
node in the graph of the model:</p>
<div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="c1"># In tensorrt_llm.functional:</span>
<span class="k">def</span> <span class="nf">activation</span><span class="p">(</span><span class="nb">input</span><span class="p">:</span> <span class="n">Tensor</span><span class="p">,</span> <span class="n">act_type</span><span class="p">:</span> <span class="n">trt</span><span class="o">.</span><span class="n">ActivationType</span><span class="p">)</span> <span class="o">-&gt;</span> <span class="n">Tensor</span><span class="p">:</span>
<span class="n">layer</span> <span class="o">=</span> <span class="n">default_trtnet</span><span class="p">()</span><span class="o">.</span><span class="n">add_activation</span><span class="p">(</span><span class="nb">input</span><span class="o">.</span><span class="n">trt_tensor</span><span class="p">,</span> <span class="n">act_type</span><span class="p">)</span> <span class="c1"># default_trtnet() -&gt; INetworkDefinition</span>
<span class="k">return</span> <span class="n">_create_tensor</span><span class="p">(</span><span class="n">layer</span><span class="o">.</span><span class="n">get_output</span><span class="p">(</span><span class="mi">0</span><span class="p">),</span> <span class="n">layer</span><span class="p">)</span>
</pre></div>
</div>
<p>To make it even easier for users, a few of the most standard activation
functions found in LLMs are derived from that function:</p>
<div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="c1"># In tensorrt_llm.functional:</span>
<span class="n">relu</span> <span class="o">=</span> <span class="n">partial</span><span class="p">(</span><span class="n">activation</span><span class="p">,</span> <span class="n">act_type</span><span class="o">=</span><span class="n">trt</span><span class="o">.</span><span class="n">ActivationType</span><span class="o">.</span><span class="n">RELU</span><span class="p">)</span>
<span class="n">sigmoid</span> <span class="o">=</span> <span class="n">partial</span><span class="p">(</span><span class="n">activation</span><span class="p">,</span> <span class="n">act_type</span><span class="o">=</span><span class="n">trt</span><span class="o">.</span><span class="n">ActivationType</span><span class="o">.</span><span class="n">SIGMOID</span><span class="p">)</span>
</pre></div>
</div>
<p>Specialized activation functions can be used to assemble more advanced
functions such as the <code class="docutils literal notranslate"><span class="pre">silu</span></code> activation:</p>
<div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="c1"># In tensorrt_llm.functional:</span>
<span class="k">def</span> <span class="nf">silu</span><span class="p">(</span><span class="nb">input</span><span class="p">:</span> <span class="n">Tensor</span><span class="p">)</span> <span class="o">-&gt;</span> <span class="n">Tensor</span><span class="p">:</span>
<span class="k">return</span> <span class="nb">input</span> <span class="o">*</span> <span class="n">sigmoid</span><span class="p">(</span><span class="nb">input</span><span class="p">)</span>
</pre></div>
</div>
<p>When the TensorRT-LLM’s Model Definition API is utilized, a graph of the network is
assembled. The graph can later be traversed or transformed using the graph
traversal API exposed by the
<a class="reference external" href="https://docs.nvidia.com/deeplearning/tensorrt/api/python_api/infer/Graph/LayerBase.html#tensorrt.ILayer"><code class="docutils literal notranslate"><span class="pre">tensorrt.ILayer</span></code></a>
class. That graph will also be optimized by TensorRT during the compilation of
the engine, as explained in the next section.</p>
</section>
<section id="compilation">
<h1>Compilation<a class="headerlink" href="#compilation" title="Link to this heading"></a></h1>
<p>Once populated, the instance of the
<a class="reference external" href="https://docs.nvidia.com/deeplearning/tensorrt/api/python_api/infer/Graph/Network.html#tensorrt.INetworkDefinition"><code class="docutils literal notranslate"><span class="pre">tensorrt.INetworkDefinition</span></code></a>,
can be compiled into an efficient engine by the
<a class="reference external" href="https://docs.nvidia.com/deeplearning/tensorrt/api/python_api/infer/Core/Builder.html#tensorrt.Builder"><code class="docutils literal notranslate"><span class="pre">tensorrt.Builder</span></code></a>
In TensorRT-LLM, it is done through the <code class="docutils literal notranslate"><span class="pre">build_engine</span></code> member function of the
<code class="docutils literal notranslate"><span class="pre">tensorrt_llm.Builder</span></code> class that calls the
<a class="reference external" href="https://docs.nvidia.com/deeplearning/tensorrt/api/python_api/infer/Core/Builder.html#tensorrt.Builder.build_serialized_network"><code class="docutils literal notranslate"><span class="pre">build_serialized_network</span></code></a>
method of the
<a class="reference external" href="https://docs.nvidia.com/deeplearning/tensorrt/api/python_api/infer/Core/Builder.html#tensorrt.Builder"><code class="docutils literal notranslate"><span class="pre">tensorrt.Builder</span></code></a>
object. That call, if everything works as expected, produces an instance of the
<a class="reference external" href="https://docs.nvidia.com/deeplearning/tensorrt/api/python_api/infer/FoundationalTypes/HostMemory.html#tensorrt.IHostMemory"><code class="docutils literal notranslate"><span class="pre">tensorrt.IHostMemory</span></code></a>
class. That object is an optimized TensorRT engine that can be stored as a
binary file.</p>
<section id="tensorrt-compiler">
<h2>TensorRT Compiler<a class="headerlink" href="#tensorrt-compiler" title="Link to this heading"></a></h2>
<p>The TensorRT compiler can sweep through the graph to choose the best kernel for each operation and available GPU. Crucially, it can also identify patterns in the graph where multiple operations are good candidates for being fused into a single kernel. This reduces the required amount of memory movement and the overhead of launching multiple GPU kernels.</p>
<p>TensorRT also compiles the graph of operations into a single <a class="reference external" href="https://developer.nvidia.com/blog/cuda-graphs/">CUDA Graph</a> that can be launched all at one time, further reducing the kernel launch overhead.</p>
<p>The TensorRT compiler is extremely powerful for fusing layers and increasing execution speed, but there are some complex layer fusions—like <a class="reference external" href="https://arxiv.org/abs/2307.08691">FlashAttention</a> — that involve interleaving many operations together and which can’t be automatically discovered. For those, you can explicitly replace parts of the graph with <a class="reference internal" href="#plugins">plugins</a> at compile time.</p>
</section>
<section id="model-engine">
<h2>Model Engine<a class="headerlink" href="#model-engine" title="Link to this heading"></a></h2>
<p>The engine file contains the information that you need for executing the model, but LLM usage in practice requires much more than a single forward pass through the model. TensorRT-LLM includes a highly optimized C++ runtime for executing built LLM engines and managing processes like sampling tokens from the model output, managing the KV cache, and batching requests together.</p>
<p>You can use that runtime directly to execute the model locally, or you can use the TensorRT-LLM runtime backend for NVIDIA Triton Inference Server to serve the model for multiple users.</p>
</section>
<section id="weight-bindings">
<h2>Weight Bindings<a class="headerlink" href="#weight-bindings" title="Link to this heading"></a></h2>
<p>TensorRT engines embed the network weights, that must be known for compilation.
For that reason, the weights must be bound to parameters in the model
definition before calling <code class="docutils literal notranslate"><span class="pre">tensorrt_llm.Builder.build_engine</span></code>. It leads to code like:</p>
<div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="c1"># The Linear operator exposes two parameters (see tensorrt_llm/layers/linear.py):</span>
<span class="k">class</span> <span class="nc">Linear</span><span class="p">(</span><span class="n">Module</span><span class="p">):</span>
<span class="k">def</span> <span class="fm">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="o">...</span><span class="p">):</span>
<span class="bp">self</span><span class="o">.</span><span class="n">weight</span> <span class="o">=</span> <span class="n">Parameter</span><span class="p">(</span><span class="n">shape</span><span class="o">=</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">out_features</span><span class="p">,</span> <span class="bp">self</span><span class="o">.</span><span class="n">in_features</span><span class="p">),</span> <span class="n">dtype</span><span class="o">=</span><span class="n">dtype</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">bias</span> <span class="o">=</span> <span class="n">Parameter</span><span class="p">(</span><span class="n">shape</span><span class="o">=</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">out_features</span><span class="p">,</span> <span class="p">),</span> <span class="n">dtype</span><span class="o">=</span><span class="n">dtype</span><span class="p">)</span>
<span class="c1"># The parameters are bound to the weights before compiling the model. See examples/gpt/weight.py:</span>
<span class="n">tensorrt_llm_gpt</span><span class="o">.</span><span class="n">layers</span><span class="p">[</span><span class="n">i</span><span class="p">]</span><span class="o">.</span><span class="n">mlp</span><span class="o">.</span><span class="n">fc</span><span class="o">.</span><span class="n">weight</span><span class="o">.</span><span class="n">value</span> <span class="o">=</span> <span class="n">fromfile</span><span class="p">(</span><span class="o">...</span><span class="p">)</span>
<span class="n">tensorrt_llm_gpt</span><span class="o">.</span><span class="n">layers</span><span class="p">[</span><span class="n">i</span><span class="p">]</span><span class="o">.</span><span class="n">mlp</span><span class="o">.</span><span class="n">fc</span><span class="o">.</span><span class="n">bias</span><span class="o">.</span><span class="n">value</span> <span class="o">=</span> <span class="n">fromfile</span><span class="p">(</span><span class="o">...</span><span class="p">)</span>
</pre></div>
</div>
<p>Note that TensorRT can also
<a class="reference external" href="https://docs.nvidia.com/deeplearning/tensorrt/developer-guide/index.html#refitting-engine-c">refit</a>
engines to update the weights after compilation. This feature is available to
TensorRT-LLM users through the <code class="docutils literal notranslate"><span class="pre">refit_engine</span></code> method in the
<code class="docutils literal notranslate"><span class="pre">tensorrt_llm.Builder</span></code> class.</p>
</section>
<section id="pattern-matching-and-fusion">
<h2>Pattern-Matching and Fusion<a class="headerlink" href="#pattern-matching-and-fusion" title="Link to this heading"></a></h2>
<p>One of the key steps performed by TensorRT when it compiles the network graph
is the fusion of operations. Fusion is a well-known technique to improve the
efficiency when executing LLMs. It helps reduce the amount of data transferred
between the memory (DRAM) and the compute cores (CUDA cores as well as Tensor
Cores located on the <a class="reference external" href="https://docs.nvidia.com/cuda/cuda-c-programming-guide/index.html#introduction">Streaming
Multiprocessors</a>
of a GPU). It also removes kernel launch overhead (each time a kernel is
launched on the GPU, there is a small additional CPU cost that is called the
launch overhead). A classical example is the fusion of the activation function
with the matrix multiplication (matmul) that usually precedes it in the
network.</p>
<p>In TensorRT-LLM, when defining the model, such a sequence can be written as:</p>
<div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="n">c</span> <span class="o">=</span> <span class="n">tensorrt_llm</span><span class="o">.</span><span class="n">functional</span><span class="o">.</span><span class="n">matmul</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">b</span><span class="p">)</span>
<span class="n">c</span> <span class="o">=</span> <span class="n">tensorrt_llm</span><span class="o">.</span><span class="n">functional</span><span class="o">.</span><span class="n">relu</span><span class="p">(</span><span class="n">c</span><span class="p">)</span>
</pre></div>
</div>
<p>During inference, if the above sequence is executed without fusion, the <code class="docutils literal notranslate"><span class="pre">c</span></code>
tensor has to be written to global memory at the end of the <code class="docutils literal notranslate"><span class="pre">matmul</span></code>, read from
that same memory in <code class="docutils literal notranslate"><span class="pre">relu</span></code> and written again after <code class="docutils literal notranslate"><span class="pre">relu</span></code>. If no other
operation uses the intermediate values between <code class="docutils literal notranslate"><span class="pre">matmul</span></code> and <code class="docutils literal notranslate"><span class="pre">relu</span></code>, it is
suboptimal. That is why, during compilation, TensorRT will identify that
pattern and automatically produce a GPU kernel that applies <code class="docutils literal notranslate"><span class="pre">relu</span></code> at the end
of <code class="docutils literal notranslate"><span class="pre">matmul</span></code> without an intermediate step through global memory. With that
optimization, the <code class="docutils literal notranslate"><span class="pre">c</span></code> tensor is written only once (after <code class="docutils literal notranslate"><span class="pre">relu</span></code>) instead of
twice, and is not read between the two operations.</p>
<p>The process of identifying the sequences of operations that can be fused is
called <em>pattern-matching</em>. TensorRT has a powerful pattern-matching algorithm
that can identify a lot of possible fusions. All the identified patterns are
converted into more efficient kernels by an advanced kernel compiler.</p>
</section>
<section id="plugins">
<h2>Plugins<a class="headerlink" href="#plugins" title="Link to this heading"></a></h2>
<p>The number of possible fusions is almost infinite and some useful fusions
involve very advanced modifications of the graph. A well-known example
is the <a class="reference external" href="https://arxiv.org/abs/2205.14135">Flash-Attention</a> technique to
optimize the <a class="reference external" href="https://arxiv.org/abs/1706.03762">Multihead-Attention</a> block
found in many LLMs. Flash-Attention requires modifications to the arithmetic
performed in the sequence <code class="docutils literal notranslate"><span class="pre">BMM-Softmax-BMM</span></code> (where <code class="docutils literal notranslate"><span class="pre">BMM</span></code> stands for Batched
Matrix-Matrix product) and the interleaving of the <code class="docutils literal notranslate"><span class="pre">for</span></code>-loops of the two
batched matrix products. That’s non-trivial and not necessarily something
you can expect a compiler to “discover” on its own (or it might require the
support for a <a class="reference external" href="https://en.wikipedia.org/wiki/Polytope_model">polyhedral
model</a>).</p>
<p>As a result, even if TensorRT has a powerful pattern-matching algorithm and
supports a lot of possible fusions, there is always the risk that it cannot
identify uncommon and/or very advanced patterns. To overcome that inevitable
limitation, TensorRT offers a powerful mechanism known as
<a class="reference external" href="https://docs.nvidia.com/deeplearning/tensorrt/api/python_api/infer/Plugin/pyPlugin.html">plugins</a>.</p>
<p>The plugins are nodes inserted in the network graph definition that map to user-defined
GPU kernels. TensorRT-LLM uses a number of such plugins. They can be found in
the <a class="reference external" href="https://github.com/NVIDIA/TensorRT-LLM/tree/rel//cpp/tensorrt_llm/plugins"><code class="docutils literal notranslate"><span class="pre">cpp/tensorrt_llm/plugins</span></code></a> directory.</p>
<p>Plugins are written in C++ and follow a well-defined interface described in the
<a class="reference external" href="https://docs.nvidia.com/deeplearning/tensorrt/developer-guide/index.html#extending">Extending TensorRT with Custom Layers</a>
section of the TensorRT
<a class="reference external" href="https://docs.nvidia.com/deeplearning/tensorrt/developer-guide/index.html">Developer Guide</a>.
When executed within a TensorRT engine, plugins trigger the execution of
their encapsulated GPU kernels. A fairly simple example of plugins is the
<a class="reference external" href="https://github.com/NVIDIA/TensorRT-LLM/tree/rel//cpp/tensorrt_llm/plugins/quantizeTensorPlugin"><code class="docutils literal notranslate"><span class="pre">QuantizeTensorPlugin</span></code></a> that
triggers a CUDA kernel in the <code class="docutils literal notranslate"><span class="pre">QuantizeTensorPlugin::enqueue</span></code> member function:</p>
<div class="highlight-cpp notranslate"><div class="highlight"><pre><span></span><span class="c1">// In cpp/tensorrt_llm/plugins/quantizeTensorPlugin/quantizeTensorPlugin.cpp:</span>
<span class="kt">int</span><span class="w"> </span><span class="nf">QuantizeTensorPlugin::enqueue</span><span class="p">(...)</span><span class="w"> </span><span class="p">{</span>
<span class="w"> </span><span class="k">if</span><span class="w"> </span><span class="p">(</span><span class="n">inputDesc</span><span class="p">[</span><span class="mi">0</span><span class="p">].</span><span class="n">type</span><span class="w"> </span><span class="o">==</span><span class="w"> </span><span class="n">DataType</span><span class="o">::</span><span class="n">kFLOAT</span><span class="p">)</span><span class="w"> </span><span class="p">{</span>
<span class="w"> </span><span class="n">invokeQuantization</span><span class="o">&lt;</span><span class="kt">float</span><span class="o">&gt;</span><span class="p">(...);</span>
<span class="w"> </span><span class="p">}</span><span class="w"> </span><span class="k">else</span><span class="w"> </span><span class="p">{</span>
<span class="w"> </span><span class="n">invokeQuantization</span><span class="o">&lt;</span><span class="n">half</span><span class="o">&gt;</span><span class="p">(...);</span>
<span class="w"> </span><span class="p">}</span>
<span class="w"> </span><span class="k">return</span><span class="w"> </span><span class="mi">0</span><span class="p">;</span>
<span class="p">}</span>
<span class="c1">// In cpp/tensorrt_llm/kernels/quantization.cu:</span>
<span class="k">template</span><span class="w"> </span><span class="o">&lt;</span><span class="k">typename</span><span class="w"> </span><span class="nc">T</span><span class="o">&gt;</span>
<span class="kt">void</span><span class="w"> </span><span class="n">invokeQuantization</span><span class="p">(...)</span><span class="w"> </span><span class="p">{</span>
<span class="w"> </span><span class="c1">// The standard &lt;&lt;&lt; &gt;&gt;&gt; construct to launch CUDA kernels</span>
<span class="w"> </span><span class="n">quantizedKernel</span><span class="o">&lt;&lt;&lt;</span><span class="n">grid</span><span class="p">,</span><span class="w"> </span><span class="n">block</span><span class="p">,</span><span class="w"> </span><span class="mi">0</span><span class="p">,</span><span class="w"> </span><span class="n">stream</span><span class="o">&gt;&gt;&gt;</span><span class="p">(...);</span>
<span class="p">}</span>
</pre></div>
</div>
<p>For more details on how TensorRT-LLM implements the GPT Attention operator, see
the <a class="reference internal" href="../advanced/gpt-attention.html"><span class="std std-doc">Multi-head, Multi-query and Group-query Attention</span></a> document.</p>
</section>
</section>
<section id="runtime">
<h1>Runtime<a class="headerlink" href="#runtime" title="Link to this heading"></a></h1>
<p>TensorRT-LLM includes an API to implement Python and C++ runtimes. The role of
the runtime components is to load the TensorRT engines and drive their
execution. Typically, for an auto-regressive model like GPT, the runtime is in
charge of loading the engine that implements both the processing of the input
sequence as well as the body of the generation loop. See the <a class="reference internal" href="../advanced/gpt-runtime.html"><span class="std std-doc">GPT C++
Runtime</span></a> document for details on the C++ Runtime.</p>
</section>
<section id="multi-gpu-and-multi-node-support">
<span id="multi-gpu-multi-node"></span><h1>Multi-GPU and Multi-Node Support<a class="headerlink" href="#multi-gpu-and-multi-node-support" title="Link to this heading"></a></h1>
<p>Even if TensorRT is designed for single-GPU systems, TensorRT-LLM adds the
support for systems with multiple GPUs and nodes. It is enabled
using TensorRT plugins that wrap communication primitives from the
<a class="reference external" href="https://developer.nvidia.com/nccl">NCCL</a> library as well as a custom
plugin that optimize the All-Reduce primitive in the presence of All-to-all
connections between GPUs (through NVSwitch in DGX systems).</p>
<p>The communication plugins can be found in
<a class="reference external" href="https://github.com/NVIDIA/TensorRT-LLM/tree/rel/cpp/tensorrt_llm/plugins/ncclPlugin">cpp/tensorrt_llm/plugins/ncclPlugin</a>
and the multi-GPU functions are exposed in the TensorRT-LLM Model Definition API
as:</p>
<div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="c1"># In tensorrt_llm/functional.py:</span>
<span class="c1"># Collectives.</span>
<span class="k">def</span> <span class="nf">allreduce</span><span class="p">(</span><span class="n">tensor</span><span class="p">:</span> <span class="n">Tensor</span><span class="p">,</span> <span class="n">group</span><span class="p">:</span> <span class="n">List</span><span class="p">[</span><span class="nb">int</span><span class="p">])</span> <span class="o">-&gt;</span> <span class="n">Tensor</span>
<span class="k">def</span> <span class="nf">allgather</span><span class="p">(</span><span class="n">tensor</span><span class="p">:</span> <span class="n">Tensor</span><span class="p">,</span> <span class="n">group</span><span class="p">:</span> <span class="n">List</span><span class="p">[</span><span class="nb">int</span><span class="p">],</span> <span class="n">gather_dim</span><span class="p">:</span> <span class="nb">int</span> <span class="o">=</span> <span class="mi">0</span><span class="p">)</span> <span class="o">-&gt;</span> <span class="n">Tensor</span>
<span class="c1"># Point-to-point communication primitives.</span>
<span class="k">def</span> <span class="nf">send</span><span class="p">(</span><span class="n">tensor</span><span class="p">:</span> <span class="n">Tensor</span><span class="p">,</span> <span class="n">tgt</span><span class="p">:</span> <span class="nb">int</span><span class="p">)</span> <span class="o">-&gt;</span> <span class="n">Tensor</span>
<span class="k">def</span> <span class="nf">recv</span><span class="p">(</span><span class="n">tensor</span><span class="p">:</span> <span class="n">Tensor</span><span class="p">,</span> <span class="n">src</span><span class="p">:</span> <span class="nb">int</span><span class="p">)</span> <span class="o">-&gt;</span> <span class="n">Tensor</span>
</pre></div>
</div>
<p>The multi-GPU support can be enabled through two different modes of model
parallelism: Tensor Parallelism and Pipeline Parallelism. The former mode
splits the different layers of a model across the GPUs. Each GPU runs the
entire network and synchronizes with its siblings when needed. The Pipeline
Parallelism distributes the different layers to the GPUs. Each GPU runs a
subset of the entire model and communications happen at the boundary of those
subsets of layers. Tensor Parallelism usually leads to more balanced executions
but requires more memory bandwidth between the GPUs. Pipeline Parallelism
reduces the need for high-bandwidth communication but may incur load-balancing
issues and may be less efficient in terms of GPU utilization.</p>
<section id="examples">
<h2>Examples<a class="headerlink" href="#examples" title="Link to this heading"></a></h2>
<p>Here are examples of Llama 3.1 70B and Llama 3.1 405B showing how to perform multi-GPU and multi-node inference in TensorRT-LLM. The example of Llama 3.1 70B performs multi-GPU inference on a single node, while the example of Llama 3.1 405B performs multi-node inference.</p>
<section id="llama-3-1-70b">
<h3>Llama 3.1 70B<a class="headerlink" href="#llama-3-1-70b" title="Link to this heading"></a></h3>
<p>The following sample commands build an engine for running the Llama 3.1 70B model with tensor parallelism (TP=4) using 4 GPUs on a single node.</p>
<div class="highlight-bash notranslate"><div class="highlight"><pre><span></span><span class="nv">folder_trt_llm</span><span class="o">=</span>../TensorRT-LLM
<span class="nv">model_dir</span><span class="o">=</span>Llama-3.1-70B
<span class="nv">ckpt_dir</span><span class="o">=</span>ckpt_llama_3.1_70b
<span class="nv">engine_dir</span><span class="o">=</span>engine_llama_3.1_70b
<span class="nv">dtype</span><span class="o">=</span>bfloat16
<span class="nv">tp_size</span><span class="o">=</span><span class="m">4</span>
<span class="nv">pp_size</span><span class="o">=</span><span class="m">1</span>
<span class="nv">kv_cache_type</span><span class="o">=</span>paged
<span class="nv">max_input_len</span><span class="o">=</span><span class="m">128</span>
<span class="nv">max_output_len</span><span class="o">=</span><span class="m">128</span>
<span class="nv">max_batch_size</span><span class="o">=</span><span class="m">4</span>
<span class="nv">workers</span><span class="o">=</span><span class="k">$((</span><span class="w"> </span><span class="nv">tp_size</span><span class="w"> </span><span class="o">*</span><span class="w"> </span><span class="nv">pp_size</span><span class="w"> </span><span class="k">))</span>
python<span class="w"> </span><span class="si">${</span><span class="nv">folder_trt_llm</span><span class="si">}</span>/examples/llama/convert_checkpoint.py<span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--output_dir<span class="w"> </span><span class="si">${</span><span class="nv">ckpt_dir</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--model_dir<span class="w"> </span><span class="si">${</span><span class="nv">model_dir</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--dtype<span class="w"> </span><span class="si">${</span><span class="nv">dtype</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--tp_size<span class="w"> </span><span class="si">${</span><span class="nv">tp_size</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--pp_size<span class="w"> </span><span class="si">${</span><span class="nv">pp_size</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--workers<span class="w"> </span><span class="si">${</span><span class="nv">workers</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--use_parallel_embedding
trtllm-build<span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--output_dir<span class="w"> </span><span class="si">${</span><span class="nv">engine_dir</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--checkpoint_dir<span class="w"> </span><span class="si">${</span><span class="nv">ckpt_dir</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--gemm_plugin<span class="w"> </span><span class="si">${</span><span class="nv">dtype</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--gpt_attention_plugin<span class="w"> </span><span class="si">${</span><span class="nv">dtype</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--kv_cache_type<span class="w"> </span><span class="si">${</span><span class="nv">kv_cache_type</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--max_input_len<span class="w"> </span><span class="si">${</span><span class="nv">max_input_len</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--max_seq_len<span class="w"> </span><span class="k">$((</span><span class="w"> </span><span class="nv">max_input_len</span><span class="w"> </span><span class="o">+</span><span class="w"> </span><span class="nv">max_output_len</span><span class="w"> </span><span class="k">))</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--max_batch_size<span class="w"> </span><span class="si">${</span><span class="nv">max_batch_size</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--workers<span class="w"> </span><span class="si">${</span><span class="nv">workers</span><span class="si">}</span>
</pre></div>
</div>
<p>The following sample commands perform inference using 4 GPUs on a single node by running <code class="docutils literal notranslate"><span class="pre">examples/run.py</span></code>.</p>
<div class="highlight-bash notranslate"><div class="highlight"><pre><span></span><span class="nv">input_text</span><span class="o">=</span><span class="s2">&quot;Born in north-east France, Soyer trained as a&quot;</span>
mpirun<span class="w"> </span>-n<span class="w"> </span><span class="k">$((</span><span class="w"> </span><span class="nv">tp_size</span><span class="w"> </span><span class="o">*</span><span class="w"> </span><span class="nv">pp_size</span><span class="w"> </span><span class="k">))</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>python<span class="w"> </span><span class="si">${</span><span class="nv">folder_trt_llm</span><span class="si">}</span>/examples/run.py<span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--engine_dir<span class="w"> </span><span class="si">${</span><span class="nv">engine_dir</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--tokenizer_dir<span class="w"> </span><span class="si">${</span><span class="nv">model_dir</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--input_text<span class="w"> </span><span class="s2">&quot;</span><span class="si">${</span><span class="nv">input_text</span><span class="si">}</span><span class="s2">&quot;</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--max_output_len<span class="w"> </span><span class="si">${</span><span class="nv">max_output_len</span><span class="si">}</span>
</pre></div>
</div>
</section>
<section id="llama-3-1-405b">
<h3>Llama 3.1 405B<a class="headerlink" href="#llama-3-1-405b" title="Link to this heading"></a></h3>
<p>The following sample commands build an engine for running the Llama 3.1 405B model with tensor parallelism (TP=16) on 2 nodes that each have 8 GPUs. Although the model runs on multiple nodes, you can build the engine on a single node.</p>
<div class="highlight-bash notranslate"><div class="highlight"><pre><span></span><span class="nv">folder_trt_llm</span><span class="o">=</span>../TensorRT-LLM
<span class="nv">model_dir</span><span class="o">=</span>Llama-3.1-405B
<span class="nv">ckpt_dir</span><span class="o">=</span>ckpt_llama_3.1_405b
<span class="nv">engine_dir</span><span class="o">=</span>engine_llama_3.1_405b
<span class="nv">dtype</span><span class="o">=</span>bfloat16
<span class="nv">tp_size</span><span class="o">=</span><span class="m">16</span>
<span class="nv">pp_size</span><span class="o">=</span><span class="m">1</span>
<span class="nv">kv_cache_type</span><span class="o">=</span>paged
<span class="nv">max_input_len</span><span class="o">=</span><span class="m">128</span>
<span class="nv">max_output_len</span><span class="o">=</span><span class="m">128</span>
<span class="nv">max_batch_size</span><span class="o">=</span><span class="m">4</span>
<span class="nv">workers</span><span class="o">=</span><span class="m">8</span>
python<span class="w"> </span><span class="si">${</span><span class="nv">folder_trt_llm</span><span class="si">}</span>/examples/llama/convert_checkpoint.py<span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--output_dir<span class="w"> </span><span class="si">${</span><span class="nv">ckpt_dir</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--model_dir<span class="w"> </span><span class="si">${</span><span class="nv">model_dir</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--dtype<span class="w"> </span><span class="si">${</span><span class="nv">dtype</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--tp_size<span class="w"> </span><span class="si">${</span><span class="nv">tp_size</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--pp_size<span class="w"> </span><span class="si">${</span><span class="nv">pp_size</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--workers<span class="w"> </span><span class="si">${</span><span class="nv">workers</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--use_parallel_embedding
trtllm-build<span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--output_dir<span class="w"> </span><span class="si">${</span><span class="nv">engine_dir</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--checkpoint_dir<span class="w"> </span><span class="si">${</span><span class="nv">ckpt_dir</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--gemm_plugin<span class="w"> </span><span class="si">${</span><span class="nv">dtype</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--gpt_attention_plugin<span class="w"> </span><span class="si">${</span><span class="nv">dtype</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--kv_cache_type<span class="w"> </span><span class="si">${</span><span class="nv">kv_cache_type</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--max_input_len<span class="w"> </span><span class="si">${</span><span class="nv">max_input_len</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--max_seq_len<span class="w"> </span><span class="k">$((</span><span class="w"> </span><span class="nv">max_input_len</span><span class="w"> </span><span class="o">+</span><span class="w"> </span><span class="nv">max_output_len</span><span class="w"> </span><span class="k">))</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--max_batch_size<span class="w"> </span><span class="si">${</span><span class="nv">max_batch_size</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--workers<span class="w"> </span><span class="si">${</span><span class="nv">workers</span><span class="si">}</span>
</pre></div>
</div>
<p>The following sample script, <code class="docutils literal notranslate"><span class="pre">launch_llama_3.1_405b.sh</span></code>, shows how to perform inference with Slurm on 2 nodes that each have 8 GPUs. If you use a different workload management software, the key concern is to run the <code class="docutils literal notranslate"><span class="pre">examples/run.py</span></code> command.</p>
<div class="highlight-bash notranslate"><div class="highlight"><pre><span></span><span class="ch">#!/bin/bash</span>
<span class="c1">#SBATCH --account account</span>
<span class="c1">#SBATCH --partition partition</span>
<span class="c1">#SBATCH --job-name job-name</span>
<span class="c1">#SBATCH --time 1:00:00</span>
<span class="c1">#SBATCH --nodes 2</span>
<span class="nv">folder_trt_llm</span><span class="o">=</span>../TensorRT-LLM
<span class="nv">engine_dir</span><span class="o">=</span>engine_llama_3.1_405b
<span class="nv">model_dir</span><span class="o">=</span>Llama-3.1-405B
<span class="nv">max_output_len</span><span class="o">=</span><span class="m">128</span>
<span class="nv">input_text</span><span class="o">=</span><span class="s2">&quot;Born in north-east France, Soyer trained as a&quot;</span>
srun<span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--ntasks-per-node<span class="w"> </span><span class="m">8</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--mpi<span class="w"> </span>pmix<span class="w"> </span><span class="se">\</span>
<span class="w"> </span>python<span class="w"> </span><span class="si">${</span><span class="nv">folder_trt_llm</span><span class="si">}</span>/examples/run.py<span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--engine_dir<span class="w"> </span><span class="si">${</span><span class="nv">engine_dir</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--tokenizer_dir<span class="w"> </span><span class="si">${</span><span class="nv">model_dir</span><span class="si">}</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--input_text<span class="w"> </span><span class="s2">&quot;</span><span class="si">${</span><span class="nv">input_text</span><span class="si">}</span><span class="s2">&quot;</span><span class="w"> </span><span class="se">\</span>
<span class="w"> </span>--max_output_len<span class="w"> </span><span class="si">${</span><span class="nv">max_output_len</span><span class="si">}</span>
</pre></div>
</div>
<p>You can perform inference by running the script on the Slurm cluster.</p>
<div class="highlight-bash notranslate"><div class="highlight"><pre><span></span>sbatch<span class="w"> </span>launch_llama_3.1_405b.sh
</pre></div>
</div>
</section>
</section>
</section>
</div>
</div>
<footer><div class="rst-footer-buttons" role="navigation" aria-label="Footer">
<a href="overview.html" class="btn btn-neutral float-left" title="TensorRT-LLM Architecture" accesskey="p" rel="prev"><span class="fa fa-arrow-circle-left" aria-hidden="true"></span> Previous</a>
<a href="checkpoint.html" class="btn btn-neutral float-right" title="TensorRT-LLM Checkpoint" accesskey="n" rel="next">Next <span class="fa fa-arrow-circle-right" aria-hidden="true"></span></a>
</div>
<hr/>
<div role="contentinfo">
<jinja2.runtime.BlockReference object at 0x7f1ac76d5ac0>
<div class="footer">
<p>
Copyright © 2024 NVIDIA Corporation
</p>
<p>
<a class="Link" href="https://www.nvidia.com/en-us/about-nvidia/privacy-policy/" target="_blank" rel="noopener"
data-cms-ai="0">Privacy Policy</a> |
<a class="Link" href="https://www.nvidia.com/en-us/about-nvidia/privacy-center/" target="_blank" rel="noopener"
data-cms-ai="0">Manage My Privacy</a> |
<a class="Link" href="https://www.nvidia.com/en-us/preferences/start/" target="_blank" rel="noopener"
data-cms-ai="0">Do Not Sell or Share My Data</a> |
<a class="Link" href="https://www.nvidia.com/en-us/about-nvidia/terms-of-service/" target="_blank"
rel="noopener" data-cms-ai="0">Terms of Service</a> |
<a class="Link" href="https://www.nvidia.com/en-us/about-nvidia/accessibility/" target="_blank" rel="noopener"
data-cms-ai="0">Accessibility</a> |
<a class="Link" href="https://www.nvidia.com/en-us/about-nvidia/company-policies/" target="_blank"
rel="noopener" data-cms-ai="0">Corporate Policies</a> |
<a class="Link" href="https://www.nvidia.com/en-us/product-security/" target="_blank" rel="noopener"
data-cms-ai="0">Product Security</a> |
<a class="Link" href="https://www.nvidia.com/en-us/contact/" target="_blank" rel="noopener"
data-cms-ai="0">Contact</a>
</p>
</div>
</div>
</footer>
</div>
</div>
</section>
</div>
<script>
jQuery(function () {
SphinxRtdTheme.Navigation.enable(true);
});
</script>
</body>
</html>
Reference in New Issue View Git Blame Copy Permalink