mirror of
https://github.com/NVIDIA/TensorRT-LLM.git
synced 2026-01-14 06:27:45 +08:00
793 lines
49 KiB
HTML
793 lines
49 KiB
HTML
|
||
|
||
<!DOCTYPE html>
|
||
|
||
|
||
<html lang="en" data-content_root="../" >
|
||
|
||
<head>
|
||
<meta charset="utf-8" />
|
||
<meta name="viewport" content="width=device-width, initial-scale=1.0" /><meta name="viewport" content="width=device-width, initial-scale=1" />
|
||
|
||
<title>KV Cache Offloading — TensorRT LLM</title>
|
||
|
||
|
||
|
||
<script data-cfasync="false">
|
||
document.documentElement.dataset.mode = localStorage.getItem("mode") || "";
|
||
document.documentElement.dataset.theme = localStorage.getItem("theme") || "";
|
||
</script>
|
||
<!--
|
||
this give us a css class that will be invisible only if js is disabled
|
||
-->
|
||
<noscript>
|
||
<style>
|
||
.pst-js-only { display: none !important; }
|
||
|
||
</style>
|
||
</noscript>
|
||
|
||
<!-- Loaded before other Sphinx assets -->
|
||
<link href="../_static/styles/theme.css?digest=8878045cc6db502f8baf" rel="stylesheet" />
|
||
<link href="../_static/styles/pydata-sphinx-theme.css?digest=8878045cc6db502f8baf" rel="stylesheet" />
|
||
|
||
<link rel="stylesheet" type="text/css" href="../_static/pygments.css?v=8f2a1f02" />
|
||
<link rel="stylesheet" type="text/css" href="../_static/styles/nvidia-sphinx-theme.css?v=933278ad" />
|
||
<link rel="stylesheet" type="text/css" href="../_static/copybutton.css?v=76b2166b" />
|
||
<link rel="stylesheet" type="text/css" href="../_static/autodoc_pydantic.css" />
|
||
<link rel="stylesheet" type="text/css" href="../_static/togglebutton.css?v=13237357" />
|
||
<link rel="stylesheet" type="text/css" href="../_static/custom.css?v=19d20f17" />
|
||
|
||
<!-- So that users can add custom icons -->
|
||
<script src="../_static/scripts/fontawesome.js?digest=8878045cc6db502f8baf"></script>
|
||
<!-- Pre-loaded scripts that we'll load fully later -->
|
||
<link rel="preload" as="script" href="../_static/scripts/bootstrap.js?digest=8878045cc6db502f8baf" />
|
||
<link rel="preload" as="script" href="../_static/scripts/pydata-sphinx-theme.js?digest=8878045cc6db502f8baf" />
|
||
|
||
|
||
|
||
<script src="../_static/documentation_options.js?v=5929fcd5"></script>
|
||
<script src="../_static/doctools.js?v=9a2dae69"></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>let toggleHintShow = 'Click to show';</script>
|
||
<script>let toggleHintHide = 'Click to hide';</script>
|
||
<script>let toggleOpenOnPrint = 'true';</script>
|
||
<script src="../_static/togglebutton.js?v=4a39c7ea"></script>
|
||
<script>var togglebuttonSelector = '.toggle, .admonition.dropdown';</script>
|
||
<script>var togglebuttonSelector = '.toggle, .admonition.dropdown';</script>
|
||
<script>DOCUMENTATION_OPTIONS.pagename = 'examples/llm_kv_cache_offloading';</script>
|
||
<script>
|
||
DOCUMENTATION_OPTIONS.theme_version = '0.16.1';
|
||
DOCUMENTATION_OPTIONS.theme_switcher_json_url = './_static/switcher.json';
|
||
DOCUMENTATION_OPTIONS.theme_switcher_version_match = '1.2.0rc4';
|
||
DOCUMENTATION_OPTIONS.show_version_warning_banner =
|
||
false;
|
||
</script>
|
||
|
||
<link rel="icon" href="../_static/favicon.png"/>
|
||
|
||
<link rel="index" title="Index" href="../genindex.html" />
|
||
<link rel="search" title="Search" href="../search.html" />
|
||
<link rel="next" title="Runtime Configuration Examples" href="llm_runtime.html" />
|
||
<link rel="prev" title="KV Cache Connector" href="llm_kv_cache_connector.html" />
|
||
|
||
|
||
<meta name="viewport" content="width=device-width, initial-scale=1"/>
|
||
<meta name="docsearch:language" content="en"/>
|
||
<meta name="docsearch:version" content="1.2.0rc4" />
|
||
|
||
|
||
</head>
|
||
|
||
|
||
|
||
<body data-bs-spy="scroll" data-bs-target=".bd-toc-nav" data-offset="180" data-bs-root-margin="0px 0px -60%" data-default-mode="">
|
||
|
||
|
||
|
||
<div id="pst-skip-link" class="skip-link d-print-none"><a href="#main-content">Skip to main content</a></div>
|
||
|
||
|
||
|
||
<div id="pst-scroll-pixel-helper"></div>
|
||
|
||
<button type="button" class="btn rounded-pill" id="pst-back-to-top">
|
||
<i class="fa-solid fa-arrow-up"></i>Back to top</button>
|
||
|
||
|
||
<dialog id="pst-search-dialog">
|
||
|
||
<form class="bd-search d-flex align-items-center"
|
||
action="../search.html"
|
||
method="get">
|
||
<i class="fa-solid fa-magnifying-glass"></i>
|
||
<input type="search"
|
||
class="form-control"
|
||
name="q"
|
||
placeholder="Search the docs ..."
|
||
aria-label="Search the docs ..."
|
||
autocomplete="off"
|
||
autocorrect="off"
|
||
autocapitalize="off"
|
||
spellcheck="false"/>
|
||
<span class="search-button__kbd-shortcut"><kbd class="kbd-shortcut__modifier">Ctrl</kbd>+<kbd>K</kbd></span>
|
||
</form>
|
||
</dialog>
|
||
|
||
<div class="pst-async-banner-revealer d-none">
|
||
<aside id="bd-header-version-warning" class="d-none d-print-none" aria-label="Version warning"></aside>
|
||
</div>
|
||
|
||
|
||
<header class="bd-header navbar navbar-expand-lg bd-navbar d-print-none">
|
||
<div class="bd-header__inner bd-page-width">
|
||
<button class="pst-navbar-icon sidebar-toggle primary-toggle" aria-label="Site navigation">
|
||
<span class="fa-solid fa-bars"></span>
|
||
</button>
|
||
|
||
|
||
<div class="col-lg-3 navbar-header-items__start">
|
||
|
||
<div class="navbar-item">
|
||
|
||
|
||
|
||
|
||
|
||
<a class="navbar-brand logo" href="../index.html">
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<img src="../_static/nvidia-logo-horiz-rgb-blk-for-screen.svg" class="logo__image only-light" alt="TensorRT LLM - Home"/>
|
||
<img src="../_static/nvidia-logo-horiz-rgb-wht-for-screen.svg" class="logo__image only-dark pst-js-only" alt="TensorRT LLM - Home"/>
|
||
|
||
|
||
<p class="title logo__title">TensorRT LLM</p>
|
||
|
||
</a></div>
|
||
|
||
</div>
|
||
|
||
<div class="col-lg-9 navbar-header-items">
|
||
|
||
<div class="me-auto navbar-header-items__center">
|
||
|
||
<div class="navbar-item">
|
||
|
||
|
||
<div class="version-switcher__container dropdown pst-js-only">
|
||
<button id="pst-version-switcher-button-2"
|
||
type="button"
|
||
class="version-switcher__button btn btn-sm dropdown-toggle"
|
||
data-bs-toggle="dropdown"
|
||
aria-haspopup="listbox"
|
||
aria-controls="pst-version-switcher-list-2"
|
||
aria-label="Version switcher list"
|
||
>
|
||
Choose version <!-- this text may get changed later by javascript -->
|
||
<span class="caret"></span>
|
||
</button>
|
||
<div id="pst-version-switcher-list-2"
|
||
class="version-switcher__menu dropdown-menu list-group-flush py-0"
|
||
role="listbox" aria-labelledby="pst-version-switcher-button-2">
|
||
<!-- dropdown will be populated by javascript on page load -->
|
||
</div>
|
||
</div></div>
|
||
|
||
</div>
|
||
|
||
|
||
<div class="navbar-header-items__end">
|
||
|
||
<div class="navbar-item navbar-persistent--container">
|
||
|
||
|
||
<button class="btn search-button-field search-button__button pst-js-only" title="Search" aria-label="Search" data-bs-placement="bottom" data-bs-toggle="tooltip">
|
||
<i class="fa-solid fa-magnifying-glass"></i>
|
||
<span class="search-button__default-text">Search</span>
|
||
<span class="search-button__kbd-shortcut"><kbd class="kbd-shortcut__modifier">Ctrl</kbd>+<kbd class="kbd-shortcut__modifier">K</kbd></span>
|
||
</button>
|
||
</div>
|
||
|
||
|
||
<div class="navbar-item">
|
||
|
||
<button class="btn btn-sm nav-link pst-navbar-icon theme-switch-button pst-js-only" aria-label="Color mode" data-bs-title="Color mode" data-bs-placement="bottom" data-bs-toggle="tooltip">
|
||
<i class="theme-switch fa-solid fa-sun fa-lg" data-mode="light" title="Light"></i>
|
||
<i class="theme-switch fa-solid fa-moon fa-lg" data-mode="dark" title="Dark"></i>
|
||
<i class="theme-switch fa-solid fa-circle-half-stroke fa-lg" data-mode="auto" title="System Settings"></i>
|
||
</button></div>
|
||
|
||
</div>
|
||
|
||
</div>
|
||
|
||
|
||
<div class="navbar-persistent--mobile">
|
||
|
||
<button class="btn search-button-field search-button__button pst-js-only" title="Search" aria-label="Search" data-bs-placement="bottom" data-bs-toggle="tooltip">
|
||
<i class="fa-solid fa-magnifying-glass"></i>
|
||
<span class="search-button__default-text">Search</span>
|
||
<span class="search-button__kbd-shortcut"><kbd class="kbd-shortcut__modifier">Ctrl</kbd>+<kbd class="kbd-shortcut__modifier">K</kbd></span>
|
||
</button>
|
||
</div>
|
||
|
||
|
||
|
||
</div>
|
||
|
||
</header>
|
||
|
||
|
||
<div class="bd-container">
|
||
<div class="bd-container__inner bd-page-width">
|
||
|
||
|
||
|
||
<dialog id="pst-primary-sidebar-modal"></dialog>
|
||
<div id="pst-primary-sidebar" class="bd-sidebar-primary bd-sidebar">
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<a class="navbar-brand logo" href="../index.html">
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<img src="../_static/nvidia-logo-horiz-rgb-blk-for-screen.svg" class="logo__image only-light" alt="TensorRT LLM - Home"/>
|
||
<img src="../_static/nvidia-logo-horiz-rgb-wht-for-screen.svg" class="logo__image only-dark pst-js-only" alt="TensorRT LLM - Home"/>
|
||
|
||
|
||
<p class="title logo__title">TensorRT LLM</p>
|
||
|
||
</a>
|
||
|
||
|
||
|
||
<div class="sidebar-header-items sidebar-primary__section">
|
||
|
||
|
||
<div class="sidebar-header-items__center">
|
||
|
||
|
||
|
||
<div class="navbar-item">
|
||
|
||
|
||
<div class="version-switcher__container dropdown pst-js-only">
|
||
<button id="pst-version-switcher-button-3"
|
||
type="button"
|
||
class="version-switcher__button btn btn-sm dropdown-toggle"
|
||
data-bs-toggle="dropdown"
|
||
aria-haspopup="listbox"
|
||
aria-controls="pst-version-switcher-list-3"
|
||
aria-label="Version switcher list"
|
||
>
|
||
Choose version <!-- this text may get changed later by javascript -->
|
||
<span class="caret"></span>
|
||
</button>
|
||
<div id="pst-version-switcher-list-3"
|
||
class="version-switcher__menu dropdown-menu list-group-flush py-0"
|
||
role="listbox" aria-labelledby="pst-version-switcher-button-3">
|
||
<!-- dropdown will be populated by javascript on page load -->
|
||
</div>
|
||
</div></div>
|
||
|
||
|
||
</div>
|
||
|
||
|
||
|
||
<div class="sidebar-header-items__end">
|
||
|
||
<div class="navbar-item">
|
||
|
||
<button class="btn btn-sm nav-link pst-navbar-icon theme-switch-button pst-js-only" aria-label="Color mode" data-bs-title="Color mode" data-bs-placement="bottom" data-bs-toggle="tooltip">
|
||
<i class="theme-switch fa-solid fa-sun fa-lg" data-mode="light" title="Light"></i>
|
||
<i class="theme-switch fa-solid fa-moon fa-lg" data-mode="dark" title="Dark"></i>
|
||
<i class="theme-switch fa-solid fa-circle-half-stroke fa-lg" data-mode="auto" title="System Settings"></i>
|
||
</button></div>
|
||
|
||
</div>
|
||
|
||
</div>
|
||
|
||
<div class="sidebar-primary-items__start sidebar-primary__section">
|
||
<div class="sidebar-primary-item">
|
||
|
||
|
||
|
||
<nav class="bd-docs-nav bd-links"
|
||
aria-label="Table of Contents">
|
||
<p class="bd-links__title" role="heading" aria-level="1">Table of Contents</p>
|
||
<div class="bd-toc-item navbar-nav"><p aria-level="2" class="caption" role="heading"><span class="caption-text">Getting Started</span></p>
|
||
<ul class="nav bd-sidenav">
|
||
<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 has-children"><a class="reference internal" href="../installation/index.html">Installation</a><details><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul>
|
||
<li class="toctree-l2"><a class="reference internal" href="../installation/containers.html">Pre-built release container images on NGC</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="../installation/linux.html">Installing on Linux via <code class="docutils literal notranslate"><span class="pre">pip</span></code></a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="../installation/build-from-source-linux.html">Building from Source Code on Linux</a></li>
|
||
</ul>
|
||
</details></li>
|
||
</ul>
|
||
<p aria-level="2" class="caption" role="heading"><span class="caption-text">Deployment Guide</span></p>
|
||
<ul class="current nav bd-sidenav">
|
||
<li class="toctree-l1 current active has-children"><a class="reference internal" href="llm_api_examples.html">LLM Examples</a><details open="open"><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul class="current">
|
||
<li class="toctree-l2"><a class="reference internal" href="llm_inference.html">Generate text</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="llm_inference_async.html">Generate text asynchronously</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="llm_inference_async_streaming.html">Generate text in streaming</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="llm_inference_distributed.html">Distributed LLM Generation</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="llm_guided_decoding.html">Generate text with guided decoding</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="llm_logits_processor.html">Control generated text using logits processor</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="llm_multilora.html">Generate text with multiple LoRA adapters</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="llm_sparse_attention.html">Sparse Attention</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="llm_speculative_decoding.html">Speculative Decoding</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="llm_kv_cache_connector.html">KV Cache Connector</a></li>
|
||
<li class="toctree-l2 current active"><a class="current reference internal" href="#">KV Cache Offloading</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="llm_runtime.html">Runtime Configuration Examples</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="llm_sampling.html">Sampling Techniques Showcase</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="llm_mgmn_llm_distributed.html">Run LLM-API with pytorch backend on Slurm</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="llm_mgmn_trtllm_bench.html">Run trtllm-bench with pytorch backend on Slurm</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="llm_mgmn_trtllm_serve.html">Run trtllm-serve with pytorch backend on Slurm</a></li>
|
||
</ul>
|
||
</details></li>
|
||
<li class="toctree-l1 has-children"><a class="reference internal" href="trtllm_serve_examples.html">Online Serving Examples</a><details><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul>
|
||
<li class="toctree-l2"><a class="reference internal" href="curl_chat_client.html">Curl Chat Client</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="curl_chat_client_for_multimodal.html">Curl Chat Client For Multimodal</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="curl_completion_client.html">Curl Completion Client</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="deepseek_r1_reasoning_parser.html">Deepseek R1 Reasoning Parser</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="genai_perf_client.html">Genai Perf Client</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="genai_perf_client_for_multimodal.html">Genai Perf Client For Multimodal</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="openai_chat_client.html">OpenAI Chat Client</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="openai_chat_client_for_multimodal.html">OpenAI Chat Client for Multimodal</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="openai_completion_client.html">OpenAI Completion Client</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="openai_completion_client_for_lora.html">Openai Completion Client For Lora</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="openai_completion_client_json_schema.html">OpenAI Completion Client with JSON Schema</a></li>
|
||
</ul>
|
||
</details></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="dynamo_k8s_example.html">Dynamo K8s Example</a></li>
|
||
<li class="toctree-l1 has-children"><a class="reference internal" href="../deployment-guide/index.html">Model Recipes</a><details><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul>
|
||
<li class="toctree-l2"><a class="reference internal" href="../deployment-guide/deployment-guide-for-deepseek-r1-on-trtllm.html">Deployment Guide for DeepSeek R1 on TensorRT LLM - Blackwell & Hopper Hardware</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="../deployment-guide/deployment-guide-for-llama3.3-70b-on-trtllm.html">Deployment Guide for Llama3.3 70B on TensorRT LLM - Blackwell & Hopper Hardware</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="../deployment-guide/deployment-guide-for-llama4-scout-on-trtllm.html">Deployment Guide for Llama4 Scout 17B on TensorRT LLM - Blackwell & Hopper Hardware</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="../deployment-guide/deployment-guide-for-gpt-oss-on-trtllm.html">Deployment Guide for GPT-OSS on TensorRT-LLM - Blackwell Hardware</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="../deployment-guide/deployment-guide-for-qwen3-next-on-trtllm.html">Deployment Guide for Qwen3 Next on TensorRT LLM - Blackwell & Hopper Hardware</a></li>
|
||
</ul>
|
||
</details></li>
|
||
</ul>
|
||
<p aria-level="2" class="caption" role="heading"><span class="caption-text">Models</span></p>
|
||
<ul class="nav bd-sidenav">
|
||
<li class="toctree-l1"><a class="reference internal" href="../models/supported-models.html">Supported Models</a></li>
|
||
|
||
<li class="toctree-l1"><a class="reference internal" href="../models/adding-new-model.html">Adding a New Model</a></li>
|
||
</ul>
|
||
<p aria-level="2" class="caption" role="heading"><span class="caption-text">CLI Reference</span></p>
|
||
<ul class="nav bd-sidenav">
|
||
<li class="toctree-l1"><a class="reference internal" href="../commands/trtllm-bench.html">trtllm-bench</a></li>
|
||
|
||
<li class="toctree-l1"><a class="reference internal" href="../commands/trtllm-eval.html">trtllm-eval</a></li>
|
||
<li class="toctree-l1 has-children"><a class="reference internal" href="../commands/trtllm-serve/index.html">trtllm-serve</a><details><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul>
|
||
<li class="toctree-l2"><a class="reference internal" href="../commands/trtllm-serve/trtllm-serve.html">trtllm-serve</a></li>
|
||
<li class="toctree-l2"><a class="reference internal" href="../commands/trtllm-serve/run-benchmark-with-trtllm-serve.html">Run benchmarking with <code class="docutils literal notranslate"><span class="pre">trtllm-serve</span></code></a></li>
|
||
</ul>
|
||
</details></li>
|
||
</ul>
|
||
<p aria-level="2" class="caption" role="heading"><span class="caption-text">API Reference</span></p>
|
||
<ul class="nav bd-sidenav">
|
||
<li class="toctree-l1"><a class="reference internal" href="../llm-api/index.html">LLM API Introduction</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../llm-api/reference.html">API Reference</a></li>
|
||
</ul>
|
||
<p aria-level="2" class="caption" role="heading"><span class="caption-text">Features</span></p>
|
||
<ul class="nav bd-sidenav">
|
||
<li class="toctree-l1"><a class="reference internal" href="../features/feature-combination-matrix.html">Feature Combination Matrix</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../features/attention.html">Multi-Head, Multi-Query, and Group-Query Attention</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../features/disagg-serving.html">Disaggregated Serving</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../features/kvcache.html">KV Cache System</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../features/long-sequence.html">Long Sequences</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../features/lora.html">LoRA (Low-Rank Adaptation)</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../features/multi-modality.html">Multimodal Support in TensorRT LLM</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../features/overlap-scheduler.html">Overlap Scheduler</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../features/paged-attention-ifb-scheduler.html">Paged Attention, IFB, and Request Scheduling</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../features/parallel-strategy.html">Parallelism in TensorRT LLM</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../features/quantization.html">Quantization</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../features/sampling.html">Sampling</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../features/additional-outputs.html">Additional Outputs</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../features/speculative-decoding.html">Speculative Decoding</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../features/checkpoint-loading.html">Checkpoint Loading</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../features/auto_deploy/auto-deploy.html">AutoDeploy (Prototype)</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../features/ray-orchestrator.html">Ray Orchestrator (Prototype)</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../features/torch_compile_and_piecewise_cuda_graph.html">Torch Compile & Piecewise CUDA Graph</a></li>
|
||
</ul>
|
||
<p aria-level="2" class="caption" role="heading"><span class="caption-text">Developer Guide</span></p>
|
||
<ul class="nav bd-sidenav">
|
||
<li class="toctree-l1"><a class="reference internal" href="../developer-guide/overview.html">Architecture Overview</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../developer-guide/perf-analysis.html">Performance Analysis</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../developer-guide/perf-benchmarking.html">TensorRT LLM Benchmarking</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../developer-guide/ci-overview.html">Continuous Integration Overview</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../developer-guide/dev-containers.html">Using Dev Containers</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../developer-guide/api-change.html">LLM API Change Guide</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../developer-guide/kv-transfer.html">Introduction to KV Cache Transmission</a></li>
|
||
</ul>
|
||
<p aria-level="2" class="caption" role="heading"><span class="caption-text">Blogs</span></p>
|
||
<ul class="nav bd-sidenav">
|
||
<li class="toctree-l1"><a class="reference internal" href="../blogs/tech_blog/blog10_ADP_Balance_Strategy.html">ADP Balance Strategy</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../blogs/tech_blog/blog11_GPT_OSS_Eagle3.html">Running GPT-OSS-120B with Eagle3 Speculative Decoding on GB200/B200 (TensorRT LLM)</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../blogs/tech_blog/blog12_Combining_Guided_Decoding_and_Speculative_Decoding.html">Combining Guided Decoding and Speculative Decoding: Making CPU and GPU Cooperate Seamlessly</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../blogs/tech_blog/blog13_Inference_Time_Compute_Implementation_in_TensorRT-LLM.html">Inference Time Compute Implementation in TensorRT LLM</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../blogs/tech_blog/blog14_Scaling_Expert_Parallelism_in_TensorRT-LLM_part3.html">Scaling Expert Parallelism in TensorRT LLM (Part 3: Pushing the Performance Boundary)</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../blogs/tech_blog/blog1_Pushing_Latency_Boundaries_Optimizing_DeepSeek-R1_Performance_on_NVIDIA_B200_GPUs.html">Pushing Latency Boundaries: Optimizing DeepSeek-R1 Performance on NVIDIA B200 GPUs</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../blogs/tech_blog/blog2_DeepSeek_R1_MTP_Implementation_and_Optimization.html">DeepSeek R1 MTP Implementation and Optimization</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../blogs/tech_blog/blog3_Optimizing_DeepSeek_R1_Throughput_on_NVIDIA_Blackwell_GPUs.html">Optimizing DeepSeek R1 Throughput on NVIDIA Blackwell GPUs: A Deep Dive for Developers</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../blogs/tech_blog/blog4_Scaling_Expert_Parallelism_in_TensorRT-LLM.html">Scaling Expert Parallelism in TensorRT LLM (Part 1: Design and Implementation of Large-scale EP)</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../blogs/tech_blog/blog5_Disaggregated_Serving_in_TensorRT-LLM.html">Disaggregated Serving in TensorRT LLM</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../blogs/tech_blog/blog6_Llama4_maverick_eagle_guide.html">How to launch Llama4 Maverick + Eagle3 TensorRT LLM server</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../blogs/tech_blog/blog7_NGram_performance_Analysis_And_Auto_Enablement.html">N-Gram Speculative Decoding in TensorRT LLM</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../blogs/tech_blog/blog8_Scaling_Expert_Parallelism_in_TensorRT-LLM_part2.html">Scaling Expert Parallelism in TensorRT LLM (Part 2: Performance Status and Optimization)</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../blogs/tech_blog/blog9_Deploying_GPT_OSS_on_TRTLLM.html">Running a High Performance GPT-OSS-120B Inference Server with TensorRT LLM</a></li>
|
||
<li class="toctree-l1"><a class="reference internal" href="../blogs/Best_perf_practice_on_DeepSeek-R1_in_TensorRT-LLM.html">How to get best performance on DeepSeek-R1 in TensorRT LLM</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/XQA-kernel.html">New XQA-kernel provides 2.4x more Llama-70B throughput within the same latency budget</a></li>
|
||
<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>
|
||
</ul>
|
||
<p aria-level="2" class="caption" role="heading"><span class="caption-text">Quick Links</span></p>
|
||
<ul class="nav bd-sidenav">
|
||
<li class="toctree-l1"><a class="reference external" href="https://github.com/NVIDIA/TensorRT-LLM/releases">Releases</a></li>
|
||
<li class="toctree-l1"><a class="reference external" href="https://github.com/NVIDIA/TensorRT-LLM">Github Code</a></li>
|
||
<li class="toctree-l1"><a class="reference external" href="https://github.com/NVIDIA/TensorRT-LLM/issues?q=is%3Aissue%20state%3Aopen%20label%3Aroadmap">Roadmap</a></li>
|
||
</ul>
|
||
<p aria-level="2" class="caption" role="heading"><span class="caption-text">Use TensorRT Engine</span></p>
|
||
<ul class="nav bd-sidenav">
|
||
<li class="toctree-l1"><a class="reference internal" href="../legacy/tensorrt_quickstart.html">LLM API with TensorRT Engine</a></li>
|
||
</ul>
|
||
</div>
|
||
</nav></div>
|
||
</div>
|
||
|
||
|
||
<div class="sidebar-primary-items__end sidebar-primary__section">
|
||
</div>
|
||
|
||
|
||
|
||
</div>
|
||
|
||
<main id="main-content" class="bd-main" role="main">
|
||
|
||
|
||
<div class="bd-content">
|
||
<div class="bd-article-container">
|
||
|
||
<div class="bd-header-article d-print-none">
|
||
<div class="header-article-items header-article__inner">
|
||
|
||
<div class="header-article-items__start">
|
||
|
||
<div class="header-article-item">
|
||
|
||
<nav aria-label="Breadcrumb" class="d-print-none">
|
||
<ul class="bd-breadcrumbs">
|
||
|
||
<li class="breadcrumb-item breadcrumb-home">
|
||
<a href="../index.html" class="nav-link" aria-label="Home">
|
||
<i class="fa-solid fa-home"></i>
|
||
</a>
|
||
</li>
|
||
|
||
<li class="breadcrumb-item"><a href="llm_api_examples.html" class="nav-link">LLM Examples</a></li>
|
||
|
||
<li class="breadcrumb-item active" aria-current="page"><span class="ellipsis">KV Cache Offloading</span></li>
|
||
</ul>
|
||
</nav>
|
||
</div>
|
||
|
||
</div>
|
||
|
||
|
||
</div>
|
||
</div>
|
||
|
||
|
||
|
||
|
||
<div id="searchbox"></div>
|
||
<article class="bd-article">
|
||
|
||
<section id="kv-cache-offloading">
|
||
<h1>KV Cache Offloading<a class="headerlink" href="#kv-cache-offloading" title="Link to this heading">#</a></h1>
|
||
<p>Source <a class="github reference external" href="https://github.com/NVIDIA/TensorRT-LLM/blob/a761585d9c15b4c1249aaf65a8f90764efa83a3c/examples/llm-api/llm_kv_cache_offloading.py">NVIDIA/TensorRT-LLM</a>.</p>
|
||
<div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="linenos"> 1</span><span class="sd">'''</span>
|
||
<span class="linenos"> 2</span><span class="sd">This script demonstrates the effectiveness of KV cache host offloading in TensorRT-LLM.</span>
|
||
<span class="linenos"> 3</span>
|
||
<span class="linenos"> 4</span><span class="sd">**Scenario:**</span>
|
||
<span class="linenos"> 5</span><span class="sd">The script simulates a scenario where the GPU's KV cache is severely limited,</span>
|
||
<span class="linenos"> 6</span><span class="sd">while multiple requests with recurring prompts (like system prompts) are processed.</span>
|
||
<span class="linenos"> 7</span>
|
||
<span class="linenos"> 8</span><span class="sd">1. **Constrained GPU Cache:** The GPU KV cache is configured to be very small,</span>
|
||
<span class="linenos"> 9</span><span class="sd"> only large enough to hold the state for a single request.</span>
|
||
<span class="linenos"> 10</span><span class="sd">2. **Alternating Prompts:** Four requests are sent sequentially (batch size of 1)</span>
|
||
<span class="linenos"> 11</span><span class="sd"> with two distinct prompts in an A, B, A, B pattern.</span>
|
||
<span class="linenos"> 12</span><span class="sd">3. **Cache Eviction:** Due to the small GPU cache, processing prompt B will</span>
|
||
<span class="linenos"> 13</span><span class="sd"> force the eviction of the cache generated for prompt A.</span>
|
||
<span class="linenos"> 14</span>
|
||
<span class="linenos"> 15</span><span class="sd">**Demonstration:**</span>
|
||
<span class="linenos"> 16</span>
|
||
<span class="linenos"> 17</span><span class="sd">* **Without Offloading (Default):**</span>
|
||
<span class="linenos"> 18</span><span class="sd"> - When the first prompt 'A' is processed, its KV cache is stored on the GPU.</span>
|
||
<span class="linenos"> 19</span><span class="sd"> - When prompt 'B' arrives, the cache manager needs space and discards the cache for 'A'.</span>
|
||
<span class="linenos"> 20</span><span class="sd"> - When prompt 'A' is sent again, its cache must be recomputed from scratch.</span>
|
||
<span class="linenos"> 21</span><span class="sd"> - **Expected Outcome:** The log will show `reused blocks: 0` and `cache hit rate: 0`.</span>
|
||
<span class="linenos"> 22</span>
|
||
<span class="linenos"> 23</span><span class="sd">* **With Offloading (`--enable_offloading`):**</span>
|
||
<span class="linenos"> 24</span><span class="sd"> - When prompt 'B' arrives, the cache for 'A' is not discarded but is instead</span>
|
||
<span class="linenos"> 25</span><span class="sd"> *offloaded* from the fast GPU VRAM to the slower (but larger) host CPU RAM.</span>
|
||
<span class="linenos"> 26</span><span class="sd"> - When prompt 'A' is sent again, its KV cache is loaded back from host RAM</span>
|
||
<span class="linenos"> 27</span><span class="sd"> to the GPU, which is significantly faster than recomputing it.</span>
|
||
<span class="linenos"> 28</span><span class="sd"> - **Expected Outcome:** The log will show positive values for `reused blocks`</span>
|
||
<span class="linenos"> 29</span><span class="sd"> and a non-zero `cache hit rate`, confirming that the cache was successfully</span>
|
||
<span class="linenos"> 30</span><span class="sd"> reused from the host.</span>
|
||
<span class="linenos"> 31</span>
|
||
<span class="linenos"> 32</span><span class="sd">**How to Run & Verify:**</span>
|
||
<span class="linenos"> 33</span>
|
||
<span class="linenos"> 34</span><span class="sd">1. **Without Offloading:**</span>
|
||
<span class="linenos"> 35</span><span class="sd"> ```bash</span>
|
||
<span class="linenos"> 36</span><span class="sd"> TLLM_LOG_LEVEL=DEBUG python llm_kv_cache_offloading.py 2>&1 | tee offloading_disabled.log</span>
|
||
<span class="linenos"> 37</span><span class="sd"> ```</span>
|
||
<span class="linenos"> 38</span><span class="sd"> (Check the log for zero reuse)</span>
|
||
<span class="linenos"> 39</span>
|
||
<span class="linenos"> 40</span><span class="sd">2. **With Offloading:**</span>
|
||
<span class="linenos"> 41</span><span class="sd"> ```bash</span>
|
||
<span class="linenos"> 42</span><span class="sd"> TLLM_LOG_LEVEL=DEBUG python llm_kv_cache_offloading.py --enable_offloading 2>&1 | tee offloading_enabled.log</span>
|
||
<span class="linenos"> 43</span><span class="sd"> ```</span>
|
||
<span class="linenos"> 44</span><span class="sd"> (Check the log for non-zero reuse)</span>
|
||
<span class="linenos"> 45</span><span class="sd">'''</span>
|
||
<span class="linenos"> 46</span>
|
||
<span class="linenos"> 47</span><span class="kn">import</span><span class="w"> </span><span class="nn">argparse</span>
|
||
<span class="linenos"> 48</span>
|
||
<span class="linenos"> 49</span><span class="kn">from</span><span class="w"> </span><span class="nn">tensorrt_llm</span><span class="w"> </span><span class="kn">import</span> <span class="n">LLM</span><span class="p">,</span> <span class="n">SamplingParams</span>
|
||
<span class="linenos"> 50</span><span class="kn">from</span><span class="w"> </span><span class="nn">tensorrt_llm.llmapi</span><span class="w"> </span><span class="kn">import</span> <span class="n">KvCacheConfig</span>
|
||
<span class="linenos"> 51</span>
|
||
<span class="linenos"> 52</span>
|
||
<span class="linenos"> 53</span><span class="k">def</span><span class="w"> </span><span class="nf">main</span><span class="p">(</span><span class="n">args</span><span class="p">):</span>
|
||
<span class="linenos"> 54</span> <span class="c1"># Define two distinct prompts to simulate different requests or system prompts.</span>
|
||
<span class="linenos"> 55</span> <span class="n">prompt_a</span> <span class="o">=</span> <span class="p">(</span>
|
||
<span class="linenos"> 56</span> <span class="s2">"Returns the per-iterations statistics computed since last call to this method. "</span>
|
||
<span class="linenos"> 57</span> <span class="s2">"Contains at most iter_stats_max_iterations iterations."</span><span class="p">)</span>
|
||
<span class="linenos"> 58</span> <span class="n">prompt_b</span> <span class="o">=</span> <span class="p">(</span><span class="s2">"Use for skipping decoding step for non generation model, "</span>
|
||
<span class="linenos"> 59</span> <span class="s2">"and return the batch_output (such as mm_embeddings)"</span><span class="p">)</span>
|
||
<span class="linenos"> 60</span>
|
||
<span class="linenos"> 61</span> <span class="c1"># Use a batch size of 1 to process requests sequentially, making the cache</span>
|
||
<span class="linenos"> 62</span> <span class="c1"># eviction and reuse cycle easy to observe.</span>
|
||
<span class="linenos"> 63</span> <span class="n">max_batch_size</span> <span class="o">=</span> <span class="mi">1</span>
|
||
<span class="linenos"> 64</span> <span class="n">max_seq_len</span> <span class="o">=</span> <span class="mi">256</span>
|
||
<span class="linenos"> 65</span>
|
||
<span class="linenos"> 66</span> <span class="c1"># --- KV Cache Configuration ---</span>
|
||
<span class="linenos"> 67</span> <span class="c1"># Set a small GPU KV cache size (in number of tokens). This is crucial for the demo,</span>
|
||
<span class="linenos"> 68</span> <span class="c1"># as it's only large enough to hold the KV cache for a single request.</span>
|
||
<span class="linenos"> 69</span> <span class="n">kv_cache_max_tokens</span> <span class="o">=</span> <span class="mi">256</span>
|
||
<span class="linenos"> 70</span> <span class="c1"># Define the size of a single cache block.</span>
|
||
<span class="linenos"> 71</span> <span class="n">kv_cache_page_size</span> <span class="o">=</span> <span class="mi">16</span>
|
||
<span class="linenos"> 72</span> <span class="c1"># Enable a 1 GB host cache if offloading is requested, otherwise disable it (size 0).</span>
|
||
<span class="linenos"> 73</span> <span class="c1"># This is the key toggle for the experiment.</span>
|
||
<span class="linenos"> 74</span> <span class="n">kv_cache_host_size</span> <span class="o">=</span> <span class="mi">1024</span><span class="o">**</span><span class="mi">3</span> <span class="k">if</span> <span class="n">args</span><span class="o">.</span><span class="n">enable_offloading</span> <span class="k">else</span> <span class="mi">0</span>
|
||
<span class="linenos"> 75</span>
|
||
<span class="linenos"> 76</span> <span class="n">sampling_params</span> <span class="o">=</span> <span class="n">SamplingParams</span><span class="p">(</span><span class="n">max_tokens</span><span class="o">=</span><span class="n">max_seq_len</span><span class="p">)</span>
|
||
<span class="linenos"> 77</span>
|
||
<span class="linenos"> 78</span> <span class="n">llm</span> <span class="o">=</span> <span class="n">LLM</span><span class="p">(</span>
|
||
<span class="linenos"> 79</span> <span class="n">model</span><span class="o">=</span><span class="s2">"Qwen/Qwen3-8B"</span><span class="p">,</span>
|
||
<span class="linenos"> 80</span> <span class="n">max_batch_size</span><span class="o">=</span><span class="n">max_batch_size</span><span class="p">,</span>
|
||
<span class="linenos"> 81</span> <span class="n">max_seq_len</span><span class="o">=</span><span class="n">max_seq_len</span><span class="p">,</span>
|
||
<span class="linenos"> 82</span> <span class="n">kv_cache_config</span><span class="o">=</span><span class="n">KvCacheConfig</span><span class="p">(</span>
|
||
<span class="linenos"> 83</span> <span class="n">enable_block_reuse</span><span class="o">=</span><span class="kc">True</span><span class="p">,</span> <span class="c1"># Enable reuse of cached blocks</span>
|
||
<span class="linenos"> 84</span> <span class="n">max_tokens</span><span class="o">=</span><span class="n">kv_cache_max_tokens</span><span class="p">,</span> <span class="c1"># Max tokens in GPU cache</span>
|
||
<span class="linenos"> 85</span> <span class="n">tokens_per_block</span><span class="o">=</span><span class="n">kv_cache_page_size</span><span class="p">,</span>
|
||
<span class="linenos"> 86</span> <span class="n">host_cache_size</span><span class="o">=</span><span class="n">kv_cache_host_size</span> <span class="c1"># Host cache size for offloading</span>
|
||
<span class="linenos"> 87</span> <span class="p">))</span>
|
||
<span class="linenos"> 88</span>
|
||
<span class="linenos"> 89</span> <span class="c1"># Process four requests sequentially using two distinct prompts (A, B, A, B).</span>
|
||
<span class="linenos"> 90</span> <span class="c1"># This pattern is designed to showcase the cache eviction and reuse behavior.</span>
|
||
<span class="linenos"> 91</span> <span class="nb">print</span><span class="p">(</span><span class="s2">"--- First Round ---"</span><span class="p">)</span>
|
||
<span class="linenos"> 92</span> <span class="c1"># 1. Process prompt A. Its cache is stored on the GPU.</span>
|
||
<span class="linenos"> 93</span> <span class="n">output_a</span> <span class="o">=</span> <span class="n">llm</span><span class="o">.</span><span class="n">generate</span><span class="p">(</span><span class="n">prompt_a</span><span class="p">,</span> <span class="n">sampling_params</span><span class="p">)</span>
|
||
<span class="linenos"> 94</span> <span class="nb">print</span><span class="p">(</span>
|
||
<span class="linenos"> 95</span> <span class="sa">f</span><span class="s2">"Prompt: </span><span class="si">{</span><span class="n">output_a</span><span class="o">.</span><span class="n">prompt</span><span class="si">!r}</span><span class="s2">, Generated text: </span><span class="si">{</span><span class="n">output_a</span><span class="o">.</span><span class="n">outputs</span><span class="p">[</span><span class="mi">0</span><span class="p">]</span><span class="o">.</span><span class="n">text</span><span class="si">!r}</span><span class="s2">"</span>
|
||
<span class="linenos"> 96</span> <span class="p">)</span>
|
||
<span class="linenos"> 97</span> <span class="c1"># 2. Process prompt B. Its cache replaces/offloads A's cache.</span>
|
||
<span class="linenos"> 98</span> <span class="n">output_b</span> <span class="o">=</span> <span class="n">llm</span><span class="o">.</span><span class="n">generate</span><span class="p">(</span><span class="n">prompt_b</span><span class="p">,</span> <span class="n">sampling_params</span><span class="p">)</span>
|
||
<span class="linenos"> 99</span> <span class="nb">print</span><span class="p">(</span>
|
||
<span class="linenos">100</span> <span class="sa">f</span><span class="s2">"Prompt: </span><span class="si">{</span><span class="n">output_b</span><span class="o">.</span><span class="n">prompt</span><span class="si">!r}</span><span class="s2">, Generated text: </span><span class="si">{</span><span class="n">output_b</span><span class="o">.</span><span class="n">outputs</span><span class="p">[</span><span class="mi">0</span><span class="p">]</span><span class="o">.</span><span class="n">text</span><span class="si">!r}</span><span class="s2">"</span>
|
||
<span class="linenos">101</span> <span class="p">)</span>
|
||
<span class="linenos">102</span>
|
||
<span class="linenos">103</span> <span class="nb">print</span><span class="p">(</span><span class="s2">"</span><span class="se">\n</span><span class="s2">--- Second Round ---"</span><span class="p">)</span>
|
||
<span class="linenos">104</span> <span class="c1"># 3. Process prompt A again.</span>
|
||
<span class="linenos">105</span> <span class="c1"># - Without offloading: Must recompute from scratch.</span>
|
||
<span class="linenos">106</span> <span class="c1"># - With offloading: Recovers cache from host RAM.</span>
|
||
<span class="linenos">107</span> <span class="n">output_a</span> <span class="o">=</span> <span class="n">llm</span><span class="o">.</span><span class="n">generate</span><span class="p">(</span><span class="n">prompt_a</span><span class="p">,</span> <span class="n">sampling_params</span><span class="p">)</span>
|
||
<span class="linenos">108</span> <span class="nb">print</span><span class="p">(</span>
|
||
<span class="linenos">109</span> <span class="sa">f</span><span class="s2">"Prompt: </span><span class="si">{</span><span class="n">output_a</span><span class="o">.</span><span class="n">prompt</span><span class="si">!r}</span><span class="s2">, Generated text: </span><span class="si">{</span><span class="n">output_a</span><span class="o">.</span><span class="n">outputs</span><span class="p">[</span><span class="mi">0</span><span class="p">]</span><span class="o">.</span><span class="n">text</span><span class="si">!r}</span><span class="s2">"</span>
|
||
<span class="linenos">110</span> <span class="p">)</span>
|
||
<span class="linenos">111</span> <span class="c1"># 4. Process prompt B again.</span>
|
||
<span class="linenos">112</span> <span class="c1"># - Without offloading: Must recompute from scratch.</span>
|
||
<span class="linenos">113</span> <span class="c1"># - With offloading: Recovers cache from host RAM.</span>
|
||
<span class="linenos">114</span> <span class="n">output_b</span> <span class="o">=</span> <span class="n">llm</span><span class="o">.</span><span class="n">generate</span><span class="p">(</span><span class="n">prompt_b</span><span class="p">,</span> <span class="n">sampling_params</span><span class="p">)</span>
|
||
<span class="linenos">115</span> <span class="nb">print</span><span class="p">(</span>
|
||
<span class="linenos">116</span> <span class="sa">f</span><span class="s2">"Prompt: </span><span class="si">{</span><span class="n">output_b</span><span class="o">.</span><span class="n">prompt</span><span class="si">!r}</span><span class="s2">, Generated text: </span><span class="si">{</span><span class="n">output_b</span><span class="o">.</span><span class="n">outputs</span><span class="p">[</span><span class="mi">0</span><span class="p">]</span><span class="o">.</span><span class="n">text</span><span class="si">!r}</span><span class="s2">"</span>
|
||
<span class="linenos">117</span> <span class="p">)</span>
|
||
<span class="linenos">118</span>
|
||
<span class="linenos">119</span> <span class="n">llm</span><span class="o">.</span><span class="n">shutdown</span><span class="p">()</span>
|
||
<span class="linenos">120</span>
|
||
<span class="linenos">121</span>
|
||
<span class="linenos">122</span><span class="k">if</span> <span class="vm">__name__</span> <span class="o">==</span> <span class="s2">"__main__"</span><span class="p">:</span>
|
||
<span class="linenos">123</span> <span class="n">parser</span> <span class="o">=</span> <span class="n">argparse</span><span class="o">.</span><span class="n">ArgumentParser</span><span class="p">(</span>
|
||
<span class="linenos">124</span> <span class="n">description</span><span class="o">=</span>
|
||
<span class="linenos">125</span> <span class="s2">"A script to demonstrate the effectiveness of KV cache host offloading."</span>
|
||
<span class="linenos">126</span> <span class="p">)</span>
|
||
<span class="linenos">127</span> <span class="n">parser</span><span class="o">.</span><span class="n">add_argument</span><span class="p">(</span><span class="s1">'--enable_offloading'</span><span class="p">,</span>
|
||
<span class="linenos">128</span> <span class="n">action</span><span class="o">=</span><span class="s1">'store_true'</span><span class="p">,</span>
|
||
<span class="linenos">129</span> <span class="n">help</span><span class="o">=</span><span class="s1">'Enable host RAM for KV cache offloading.'</span><span class="p">)</span>
|
||
<span class="linenos">130</span> <span class="n">args</span> <span class="o">=</span> <span class="n">parser</span><span class="o">.</span><span class="n">parse_args</span><span class="p">()</span>
|
||
<span class="linenos">131</span> <span class="n">main</span><span class="p">(</span><span class="n">args</span><span class="p">)</span>
|
||
</pre></div>
|
||
</div>
|
||
</section>
|
||
|
||
|
||
</article>
|
||
|
||
|
||
|
||
|
||
|
||
<footer class="prev-next-footer d-print-none">
|
||
|
||
<div class="prev-next-area">
|
||
<a class="left-prev"
|
||
href="llm_kv_cache_connector.html"
|
||
title="previous page">
|
||
<i class="fa-solid fa-angle-left"></i>
|
||
<div class="prev-next-info">
|
||
<p class="prev-next-subtitle">previous</p>
|
||
<p class="prev-next-title">KV Cache Connector</p>
|
||
</div>
|
||
</a>
|
||
<a class="right-next"
|
||
href="llm_runtime.html"
|
||
title="next page">
|
||
<div class="prev-next-info">
|
||
<p class="prev-next-subtitle">next</p>
|
||
<p class="prev-next-title">Runtime Configuration Examples</p>
|
||
</div>
|
||
<i class="fa-solid fa-angle-right"></i>
|
||
</a>
|
||
</div>
|
||
</footer>
|
||
|
||
</div>
|
||
|
||
|
||
|
||
<div class="bd-sidebar-secondary"></div>
|
||
|
||
|
||
|
||
|
||
|
||
</div>
|
||
<footer class="bd-footer-content">
|
||
|
||
</footer>
|
||
|
||
</main>
|
||
</div>
|
||
</div>
|
||
|
||
|
||
<!-- Scripts loaded after <body> so the DOM is not blocked -->
|
||
<script defer src="../_static/scripts/bootstrap.js?digest=8878045cc6db502f8baf"></script>
|
||
<script defer src="../_static/scripts/pydata-sphinx-theme.js?digest=8878045cc6db502f8baf"></script>
|
||
|
||
|
||
<footer class="bd-footer">
|
||
<div class="bd-footer__inner bd-page-width">
|
||
|
||
<div class="footer-items__start">
|
||
|
||
<div class="footer-item">
|
||
<a class="footer-brand logo" href="https://www.nvidia.com">
|
||
<img src="../_static/nvidia-logo-horiz-rgb-1c-blk-for-screen.svg" class="logo__image only-light" alt="NVIDIA"/>
|
||
<img src="../_static/nvidia-logo-horiz-rgb-1c-wht-for-screen.svg" class="logo__image only-dark" alt="NVIDIA"/>
|
||
</a></div>
|
||
|
||
<div class="footer-item">
|
||
|
||
<div class="footer-links">
|
||
|
||
|
||
<a class="external" href="https://www.nvidia.com/en-us/about-nvidia/privacy-policy/">Privacy Policy</a>
|
||
|
|
||
|
||
|
||
|
||
<a class="external" href="https://www.nvidia.com/en-us/about-nvidia/privacy-center/">Your Privacy Choices</a>
|
||
|
|
||
|
||
|
||
|
||
<a class="external" href="https://www.nvidia.com/en-us/about-nvidia/terms-of-service/">Terms of Service</a>
|
||
|
|
||
|
||
|
||
|
||
<a class="external" href="https://www.nvidia.com/en-us/about-nvidia/accessibility/">Accessibility</a>
|
||
|
|
||
|
||
|
||
|
||
<a class="external" href="https://www.nvidia.com/en-us/about-nvidia/company-policies/">Corporate Policies</a>
|
||
|
|
||
|
||
|
||
|
||
<a class="external" href="https://www.nvidia.com/en-us/product-security/">Product Security</a>
|
||
|
|
||
|
||
|
||
|
||
<a class="external" href="https://www.nvidia.com/en-us/contact/">Contact</a>
|
||
|
||
|
||
|
||
</div>
|
||
</div>
|
||
|
||
<div class="footer-item">
|
||
|
||
|
||
|
||
|
||
<p class="copyright">
|
||
|
||
Copyright © 2025, NVidia.
|
||
<br/>
|
||
|
||
</p>
|
||
</div>
|
||
|
||
<div class="footer-item">
|
||
<div class="extra_footer">
|
||
|
||
<p>Last updated on November 23, 2025.</p>
|
||
|
||
<p>This page is generated by TensorRT-LLM commit <a href="https://github.com/NVIDIA/TensorRT-LLM/tree/a761585">a761585</a>.</p>
|
||
|
||
</div></div>
|
||
|
||
</div>
|
||
|
||
|
||
|
||
</div>
|
||
|
||
</footer>
|
||
</body>
|
||
</html> |