Built site for gh-pages
This commit is contained in:
Quarto GHA Workflow Runner
@@ -416,6 +416,7 @@ ul.task-list li input[type="checkbox"] {
|
||||
<li><a href="#core" id="toc-core" class="nav-link" data-scroll-target="#core">Core</a></li>
|
||||
<li><a href="#cli" id="toc-cli" class="nav-link" data-scroll-target="#cli">CLI</a></li>
|
||||
<li><a href="#trainers" id="toc-trainers" class="nav-link" data-scroll-target="#trainers">Trainers</a></li>
|
||||
<li><a href="#model-loading" id="toc-model-loading" class="nav-link" data-scroll-target="#model-loading">Model Loading</a></li>
|
||||
<li><a href="#mixins" id="toc-mixins" class="nav-link" data-scroll-target="#mixins">Mixins</a></li>
|
||||
<li><a href="#context-managers" id="toc-context-managers" class="nav-link" data-scroll-target="#context-managers">Context Managers</a></li>
|
||||
<li><a href="#prompt-strategies" id="toc-prompt-strategies" class="nav-link" data-scroll-target="#prompt-strategies">Prompt Strategies</a></li>
|
||||
@@ -612,6 +613,38 @@ ul.task-list li input[type="checkbox"] {
|
||||
</tbody>
|
||||
</table>
|
||||
</section>
|
||||
<section id="model-loading" class="level2">
|
||||
<h2 class="anchored" data-anchor-id="model-loading">Model Loading</h2>
|
||||
<p>Functionality for loading and patching models, tokenizers, etc.</p>
|
||||
<table class="caption-top table">
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td><a href="../../docs/api/loaders.model.html#axolotl.loaders.model">loaders.model</a></td>
|
||||
<td>Model loader class implementation for loading, configuring, and patching various</td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td><a href="../../docs/api/loaders.tokenizer.html#axolotl.loaders.tokenizer">loaders.tokenizer</a></td>
|
||||
<td>Tokenizer loading functionality and associated utils</td>
|
||||
</tr>
|
||||
<tr class="odd">
|
||||
<td><a href="../../docs/api/loaders.processor.html#axolotl.loaders.processor">loaders.processor</a></td>
|
||||
<td>Processor loading functionality for multi-modal models</td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td><a href="../../docs/api/loaders.adapter.html#axolotl.loaders.adapter">loaders.adapter</a></td>
|
||||
<td>Adapter loading functionality, including LoRA / QLoRA and associated utils</td>
|
||||
</tr>
|
||||
<tr class="odd">
|
||||
<td><a href="../../docs/api/loaders.patch_manager.html#axolotl.loaders.patch_manager">loaders.patch_manager</a></td>
|
||||
<td>Patch manager class implementation to complement <code>axolotl.loaders.ModelLoader</code>.</td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td><a href="../../docs/api/loaders.constants.html#axolotl.loaders.constants">loaders.constants</a></td>
|
||||
<td>Shared constants for axolotl.loaders module</td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
</section>
|
||||
<section id="mixins" class="level2">
|
||||
<h2 class="anchored" data-anchor-id="mixins">Mixins</h2>
|
||||
<p>Mixin classes for augmenting trainers</p>
|
||||
@@ -853,6 +886,14 @@ ul.task-list li input[type="checkbox"] {
|
||||
<td><a href="../../docs/api/monkeypatch.mixtral.html#axolotl.monkeypatch.mixtral">monkeypatch.mixtral</a></td>
|
||||
<td>Patches to support multipack for mixtral</td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td><a href="../../docs/api/monkeypatch.gradient_checkpointing.offload_cpu.html#axolotl.monkeypatch.gradient_checkpointing.offload_cpu">monkeypatch.gradient_checkpointing.offload_cpu</a></td>
|
||||
<td>CPU offloaded checkpointing</td>
|
||||
</tr>
|
||||
<tr class="odd">
|
||||
<td><a href="../../docs/api/monkeypatch.gradient_checkpointing.offload_disk.html#axolotl.monkeypatch.gradient_checkpointing.offload_disk">monkeypatch.gradient_checkpointing.offload_disk</a></td>
|
||||
<td>DISCO - DIsk-based Storage and Checkpointing with Optimized prefetching</td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
</section>
|
||||
@@ -862,25 +903,17 @@ ul.task-list li input[type="checkbox"] {
|
||||
<table class="caption-top table">
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td><a href="../../docs/api/utils.models.html#axolotl.utils.models">utils.models</a></td>
|
||||
<td>Module for models and model loading</td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td><a href="../../docs/api/utils.tokenization.html#axolotl.utils.tokenization">utils.tokenization</a></td>
|
||||
<td>Module for tokenization utilities</td>
|
||||
</tr>
|
||||
<tr class="odd">
|
||||
<tr class="even">
|
||||
<td><a href="../../docs/api/utils.chat_templates.html#axolotl.utils.chat_templates">utils.chat_templates</a></td>
|
||||
<td>This module provides functionality for selecting chat templates based on user choices.</td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<tr class="odd">
|
||||
<td><a href="../../docs/api/utils.lora.html#axolotl.utils.lora">utils.lora</a></td>
|
||||
<td>module to get the state dict of a merged lora model</td>
|
||||
</tr>
|
||||
<tr class="odd">
|
||||
<td><a href="../../docs/api/utils.lora_embeddings.html#axolotl.utils.lora_embeddings">utils.lora_embeddings</a></td>
|
||||
<td>helpers for lora embeddings</td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td><a href="../../docs/api/utils.model_shard_quant.html#axolotl.utils.model_shard_quant">utils.model_shard_quant</a></td>
|
||||
<td>module to handle loading model on cpu/meta device for FSDP</td>
|
||||
@@ -921,14 +954,6 @@ ul.task-list li input[type="checkbox"] {
|
||||
<td><a href="../../docs/api/utils.data.sft.html#axolotl.utils.data.sft">utils.data.sft</a></td>
|
||||
<td>data handling specific to SFT</td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td><a href="../../docs/api/utils.gradient_checkpointing.offload_cpu.html#axolotl.utils.gradient_checkpointing.offload_cpu">utils.gradient_checkpointing.offload_cpu</a></td>
|
||||
<td>CPU offloaded checkpointing</td>
|
||||
</tr>
|
||||
<tr class="odd">
|
||||
<td><a href="../../docs/api/utils.gradient_checkpointing.offload_disk.html#axolotl.utils.gradient_checkpointing.offload_disk">utils.gradient_checkpointing.offload_disk</a></td>
|
||||
<td>DISCO - DIsk-based Storage and Checkpointing with Optimized prefetching</td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
</section>
|
||||
|
||||
File diff suppressed because it is too large
Load Diff
@@ -7,7 +7,7 @@
|
||||
<meta name="viewport" content="width=device-width, initial-scale=1.0, user-scalable=yes">
|
||||
|
||||
|
||||
<title>utils.lora_embeddings – Axolotl</title>
|
||||
<title>loaders.adapter – Axolotl</title>
|
||||
<style>
|
||||
code{white-space: pre-wrap;}
|
||||
span.smallcaps{font-variant: small-caps;}
|
||||
@@ -446,11 +446,12 @@ pre > code.sourceCode > span > a:first-child::before { text-decoration: underlin
|
||||
<h2 id="toc-title">On this page</h2>
|
||||
|
||||
<ul>
|
||||
<li><a href="#axolotl.utils.lora_embeddings" id="toc-axolotl.utils.lora_embeddings" class="nav-link active" data-scroll-target="#axolotl.utils.lora_embeddings">utils.lora_embeddings</a>
|
||||
<li><a href="#axolotl.loaders.adapter" id="toc-axolotl.loaders.adapter" class="nav-link active" data-scroll-target="#axolotl.loaders.adapter">loaders.adapter</a>
|
||||
<ul class="collapse">
|
||||
<li><a href="#functions" id="toc-functions" class="nav-link" data-scroll-target="#functions">Functions</a>
|
||||
<ul class="collapse">
|
||||
<li><a href="#axolotl.utils.lora_embeddings.get_linear_embedding_layers" id="toc-axolotl.utils.lora_embeddings.get_linear_embedding_layers" class="nav-link" data-scroll-target="#axolotl.utils.lora_embeddings.get_linear_embedding_layers">get_linear_embedding_layers</a></li>
|
||||
<li><a href="#axolotl.loaders.adapter.setup_quantized_meta_for_peft" id="toc-axolotl.loaders.adapter.setup_quantized_meta_for_peft" class="nav-link" data-scroll-target="#axolotl.loaders.adapter.setup_quantized_meta_for_peft">setup_quantized_meta_for_peft</a></li>
|
||||
<li><a href="#axolotl.loaders.adapter.setup_quantized_peft_meta_for_training" id="toc-axolotl.loaders.adapter.setup_quantized_peft_meta_for_training" class="nav-link" data-scroll-target="#axolotl.loaders.adapter.setup_quantized_peft_meta_for_training">setup_quantized_peft_meta_for_training</a></li>
|
||||
</ul></li>
|
||||
</ul></li>
|
||||
</ul>
|
||||
@@ -462,10 +463,10 @@ pre > code.sourceCode > span > a:first-child::before { text-decoration: underlin
|
||||
|
||||
|
||||
|
||||
<section id="axolotl.utils.lora_embeddings" class="level1">
|
||||
<h1>utils.lora_embeddings</h1>
|
||||
<p><code>utils.lora_embeddings</code></p>
|
||||
<p>helpers for lora embeddings</p>
|
||||
<section id="axolotl.loaders.adapter" class="level1">
|
||||
<h1>loaders.adapter</h1>
|
||||
<p><code>loaders.adapter</code></p>
|
||||
<p>Adapter loading functionality, including LoRA / QLoRA and associated utils</p>
|
||||
<section id="functions" class="level2">
|
||||
<h2 class="anchored" data-anchor-id="functions">Functions</h2>
|
||||
<table class="caption-top table">
|
||||
@@ -477,15 +478,24 @@ pre > code.sourceCode > span > a:first-child::before { text-decoration: underlin
|
||||
</thead>
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td><a href="#axolotl.utils.lora_embeddings.get_linear_embedding_layers">get_linear_embedding_layers</a></td>
|
||||
<td>returns the linear embedding layers needed for loras, dependent on the model arch</td>
|
||||
<td><a href="#axolotl.loaders.adapter.setup_quantized_meta_for_peft">setup_quantized_meta_for_peft</a></td>
|
||||
<td>Replaces <code>quant_state.to</code> with a dummy function to prevent PEFT from moving <code>quant_state</code> to meta device</td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td><a href="#axolotl.loaders.adapter.setup_quantized_peft_meta_for_training">setup_quantized_peft_meta_for_training</a></td>
|
||||
<td>Replaces dummy <code>quant_state.to</code> method with the original function to allow training to continue</td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
<section id="axolotl.utils.lora_embeddings.get_linear_embedding_layers" class="level3">
|
||||
<h3 class="anchored" data-anchor-id="axolotl.utils.lora_embeddings.get_linear_embedding_layers">get_linear_embedding_layers</h3>
|
||||
<div class="sourceCode" id="cb1"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb1-1"><a href="#cb1-1" aria-hidden="true" tabindex="-1"></a>utils.lora_embeddings.get_linear_embedding_layers(model_type)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>returns the linear embedding layers needed for loras, dependent on the model arch</p>
|
||||
<section id="axolotl.loaders.adapter.setup_quantized_meta_for_peft" class="level3">
|
||||
<h3 class="anchored" data-anchor-id="axolotl.loaders.adapter.setup_quantized_meta_for_peft">setup_quantized_meta_for_peft</h3>
|
||||
<div class="sourceCode" id="cb1"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb1-1"><a href="#cb1-1" aria-hidden="true" tabindex="-1"></a>loaders.adapter.setup_quantized_meta_for_peft(model)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Replaces <code>quant_state.to</code> with a dummy function to prevent PEFT from moving <code>quant_state</code> to meta device</p>
|
||||
</section>
|
||||
<section id="axolotl.loaders.adapter.setup_quantized_peft_meta_for_training" class="level3">
|
||||
<h3 class="anchored" data-anchor-id="axolotl.loaders.adapter.setup_quantized_peft_meta_for_training">setup_quantized_peft_meta_for_training</h3>
|
||||
<div class="sourceCode" id="cb2"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb2-1"><a href="#cb2-1" aria-hidden="true" tabindex="-1"></a>loaders.adapter.setup_quantized_peft_meta_for_training(model)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Replaces dummy <code>quant_state.to</code> method with the original function to allow training to continue</p>
|
||||
|
||||
|
||||
</section>
|
||||
837
docs/api/loaders.constants.html
Normal file
837
docs/api/loaders.constants.html
Normal file
@@ -0,0 +1,837 @@
|
||||
<!DOCTYPE html>
|
||||
<html xmlns="http://www.w3.org/1999/xhtml" lang="en" xml:lang="en"><head>
|
||||
|
||||
<meta charset="utf-8">
|
||||
<meta name="generator" content="quarto-1.7.31">
|
||||
|
||||
<meta name="viewport" content="width=device-width, initial-scale=1.0, user-scalable=yes">
|
||||
|
||||
|
||||
<title>loaders.constants – Axolotl</title>
|
||||
<style>
|
||||
code{white-space: pre-wrap;}
|
||||
span.smallcaps{font-variant: small-caps;}
|
||||
div.columns{display: flex; gap: min(4vw, 1.5em);}
|
||||
div.column{flex: auto; overflow-x: auto;}
|
||||
div.hanging-indent{margin-left: 1.5em; text-indent: -1.5em;}
|
||||
ul.task-list{list-style: none;}
|
||||
ul.task-list li input[type="checkbox"] {
|
||||
width: 0.8em;
|
||||
margin: 0 0.8em 0.2em -1em; /* quarto-specific, see https://github.com/quarto-dev/quarto-cli/issues/4556 */
|
||||
vertical-align: middle;
|
||||
}
|
||||
</style>
|
||||
|
||||
|
||||
<script src="../../site_libs/quarto-nav/quarto-nav.js"></script>
|
||||
<script src="../../site_libs/clipboard/clipboard.min.js"></script>
|
||||
<script src="../../site_libs/quarto-search/autocomplete.umd.js"></script>
|
||||
<script src="../../site_libs/quarto-search/fuse.min.js"></script>
|
||||
<script src="../../site_libs/quarto-search/quarto-search.js"></script>
|
||||
<meta name="quarto:offset" content="../../">
|
||||
<link href="../../favicon.jpg" rel="icon" type="image/jpeg">
|
||||
<script src="../../site_libs/quarto-html/quarto.js" type="module"></script>
|
||||
<script src="../../site_libs/quarto-html/tabsets/tabsets.js" type="module"></script>
|
||||
<script src="../../site_libs/quarto-html/popper.min.js"></script>
|
||||
<script src="../../site_libs/quarto-html/tippy.umd.min.js"></script>
|
||||
<script src="../../site_libs/quarto-html/anchor.min.js"></script>
|
||||
<link href="../../site_libs/quarto-html/tippy.css" rel="stylesheet">
|
||||
<link href="../../site_libs/quarto-html/quarto-syntax-highlighting-dark-8ef56b68f8fa1e9d2ba328e99e439f80.css" rel="stylesheet" id="quarto-text-highlighting-styles">
|
||||
<script src="../../site_libs/bootstrap/bootstrap.min.js"></script>
|
||||
<link href="../../site_libs/bootstrap/bootstrap-icons.css" rel="stylesheet">
|
||||
<link href="../../site_libs/bootstrap/bootstrap-2288ecdcbf81d2ab6432743cedd71d9a.min.css" rel="stylesheet" append-hash="true" id="quarto-bootstrap" data-mode="dark">
|
||||
<script id="quarto-search-options" type="application/json">{
|
||||
"location": "navbar",
|
||||
"copy-button": false,
|
||||
"collapse-after": 3,
|
||||
"panel-placement": "end",
|
||||
"type": "overlay",
|
||||
"limit": 50,
|
||||
"keyboard-shortcut": [
|
||||
"f",
|
||||
"/",
|
||||
"s"
|
||||
],
|
||||
"show-item-context": false,
|
||||
"language": {
|
||||
"search-no-results-text": "No results",
|
||||
"search-matching-documents-text": "matching documents",
|
||||
"search-copy-link-title": "Copy link to search",
|
||||
"search-hide-matches-text": "Hide additional matches",
|
||||
"search-more-match-text": "more match in this document",
|
||||
"search-more-matches-text": "more matches in this document",
|
||||
"search-clear-button-title": "Clear",
|
||||
"search-text-placeholder": "",
|
||||
"search-detached-cancel-button-title": "Cancel",
|
||||
"search-submit-button-title": "Submit",
|
||||
"search-label": "Search"
|
||||
}
|
||||
}</script>
|
||||
|
||||
|
||||
<link rel="stylesheet" href="../../styles.css">
|
||||
</head>
|
||||
|
||||
<body class="nav-sidebar docked nav-fixed quarto-light">
|
||||
|
||||
<div id="quarto-search-results"></div>
|
||||
<header id="quarto-header" class="headroom fixed-top">
|
||||
<nav class="navbar navbar-expand " data-bs-theme="dark">
|
||||
<div class="navbar-container container-fluid">
|
||||
<div class="navbar-brand-container mx-auto">
|
||||
<a href="../../index.html" class="navbar-brand navbar-brand-logo">
|
||||
<img src="../../image/axolotl_logo_digital_white.svg" alt="" class="navbar-logo">
|
||||
</a>
|
||||
</div>
|
||||
<div class="quarto-navbar-tools tools-wide tools-end">
|
||||
<a href="https://twitter.com/axolotl_ai" title="" class="quarto-navigation-tool px-1" aria-label=""><i class="bi bi-twitter"></i></a>
|
||||
<a href="https://github.com/axolotl-ai-cloud/axolotl/" title="" class="quarto-navigation-tool px-1" aria-label=""><i class="bi bi-github"></i></a>
|
||||
<a href="https://discord.gg/7m9sfhzaf3" title="" class="quarto-navigation-tool px-1" aria-label=""><i class="bi bi-discord"></i></a>
|
||||
</div>
|
||||
<div id="quarto-search" class="" title="Search"></div>
|
||||
</div> <!-- /container-fluid -->
|
||||
</nav>
|
||||
<nav class="quarto-secondary-nav">
|
||||
<div class="container-fluid d-flex">
|
||||
<button type="button" class="quarto-btn-toggle btn" data-bs-toggle="collapse" role="button" data-bs-target=".quarto-sidebar-collapse-item" aria-controls="quarto-sidebar" aria-expanded="false" aria-label="Toggle sidebar navigation" onclick="if (window.quartoToggleHeadroom) { window.quartoToggleHeadroom(); }">
|
||||
<i class="bi bi-layout-text-sidebar-reverse"></i>
|
||||
</button>
|
||||
<nav class="quarto-page-breadcrumbs" aria-label="breadcrumb"><ol class="breadcrumb"></ol></nav>
|
||||
<a class="flex-grow-1" role="navigation" data-bs-toggle="collapse" data-bs-target=".quarto-sidebar-collapse-item" aria-controls="quarto-sidebar" aria-expanded="false" aria-label="Toggle sidebar navigation" onclick="if (window.quartoToggleHeadroom) { window.quartoToggleHeadroom(); }">
|
||||
</a>
|
||||
</div>
|
||||
</nav>
|
||||
</header>
|
||||
<!-- content -->
|
||||
<div id="quarto-content" class="quarto-container page-columns page-rows-contents page-layout-article page-navbar">
|
||||
<!-- sidebar -->
|
||||
<nav id="quarto-sidebar" class="sidebar collapse collapse-horizontal quarto-sidebar-collapse-item sidebar-navigation docked overflow-auto">
|
||||
<div class="sidebar-menu-container">
|
||||
<ul class="list-unstyled mt-1">
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../index.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Home</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item sidebar-item-section">
|
||||
<div class="sidebar-item-container">
|
||||
<a class="sidebar-item-text sidebar-link text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-1" role="navigation" aria-expanded="true">
|
||||
<span class="menu-text">Getting Started</span></a>
|
||||
<a class="sidebar-item-toggle text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-1" role="navigation" aria-expanded="true" aria-label="Toggle section">
|
||||
<i class="bi bi-chevron-right ms-2"></i>
|
||||
</a>
|
||||
</div>
|
||||
<ul id="quarto-sidebar-section-1" class="collapse list-unstyled sidebar-section depth1 show">
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/getting-started.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Quickstart</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/installation.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Installation</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/inference.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Inference and Merging</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/cli.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Command Line Interface (CLI)</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/config.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Config Reference</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/api" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">API Reference</span></a>
|
||||
</div>
|
||||
</li>
|
||||
</ul>
|
||||
</li>
|
||||
<li class="sidebar-item sidebar-item-section">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/dataset-formats/index.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Dataset Formats</span></a>
|
||||
<a class="sidebar-item-toggle text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-2" role="navigation" aria-expanded="true" aria-label="Toggle section">
|
||||
<i class="bi bi-chevron-right ms-2"></i>
|
||||
</a>
|
||||
</div>
|
||||
<ul id="quarto-sidebar-section-2" class="collapse list-unstyled sidebar-section depth1 show">
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/dataset-formats/pretraining.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Pre-training</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/dataset-formats/inst_tune.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Instruction Tuning</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/dataset-formats/conversation.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Conversation</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/dataset-formats/stepwise_supervised.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Stepwise Supervised Format</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/dataset-formats/template_free.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Template-Free</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/dataset-formats/tokenized.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Custom Pre-Tokenized Dataset</span></a>
|
||||
</div>
|
||||
</li>
|
||||
</ul>
|
||||
</li>
|
||||
<li class="sidebar-item sidebar-item-section">
|
||||
<div class="sidebar-item-container">
|
||||
<a class="sidebar-item-text sidebar-link text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-3" role="navigation" aria-expanded="true">
|
||||
<span class="menu-text">Deployments</span></a>
|
||||
<a class="sidebar-item-toggle text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-3" role="navigation" aria-expanded="true" aria-label="Toggle section">
|
||||
<i class="bi bi-chevron-right ms-2"></i>
|
||||
</a>
|
||||
</div>
|
||||
<ul id="quarto-sidebar-section-3" class="collapse list-unstyled sidebar-section depth1 show">
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/docker.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Docker</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/multi-gpu.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Multi-GPU</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/multi-node.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Multi Node</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/ray-integration.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Ray Train</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/amd_hpc.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">AMD GPUs on HPC Systems</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/mac.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Mac M-series</span></a>
|
||||
</div>
|
||||
</li>
|
||||
</ul>
|
||||
</li>
|
||||
<li class="sidebar-item sidebar-item-section">
|
||||
<div class="sidebar-item-container">
|
||||
<a class="sidebar-item-text sidebar-link text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-4" role="navigation" aria-expanded="true">
|
||||
<span class="menu-text">How To Guides</span></a>
|
||||
<a class="sidebar-item-toggle text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-4" role="navigation" aria-expanded="true" aria-label="Toggle section">
|
||||
<i class="bi bi-chevron-right ms-2"></i>
|
||||
</a>
|
||||
</div>
|
||||
<ul id="quarto-sidebar-section-4" class="collapse list-unstyled sidebar-section depth1 show">
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/multimodal.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">MultiModal / Vision Language Models (BETA)</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/rlhf.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">RLHF (Beta)</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/reward_modelling.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Reward Modelling</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/lr_groups.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Learning Rate Groups</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/lora_optims.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">LoRA Optimizations</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/dataset_loading.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Dataset Loading</span></a>
|
||||
</div>
|
||||
</li>
|
||||
</ul>
|
||||
</li>
|
||||
<li class="sidebar-item sidebar-item-section">
|
||||
<div class="sidebar-item-container">
|
||||
<a class="sidebar-item-text sidebar-link text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-5" role="navigation" aria-expanded="true">
|
||||
<span class="menu-text">Core Concepts</span></a>
|
||||
<a class="sidebar-item-toggle text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-5" role="navigation" aria-expanded="true" aria-label="Toggle section">
|
||||
<i class="bi bi-chevron-right ms-2"></i>
|
||||
</a>
|
||||
</div>
|
||||
<ul id="quarto-sidebar-section-5" class="collapse list-unstyled sidebar-section depth1 show">
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/batch_vs_grad.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Batch size vs Gradient accumulation</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/dataset_preprocessing.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Dataset Preprocessing</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/multipack.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Multipack (Sample Packing)</span></a>
|
||||
</div>
|
||||
</li>
|
||||
</ul>
|
||||
</li>
|
||||
<li class="sidebar-item sidebar-item-section">
|
||||
<div class="sidebar-item-container">
|
||||
<a class="sidebar-item-text sidebar-link text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-6" role="navigation" aria-expanded="true">
|
||||
<span class="menu-text">Advanced Features</span></a>
|
||||
<a class="sidebar-item-toggle text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-6" role="navigation" aria-expanded="true" aria-label="Toggle section">
|
||||
<i class="bi bi-chevron-right ms-2"></i>
|
||||
</a>
|
||||
</div>
|
||||
<ul id="quarto-sidebar-section-6" class="collapse list-unstyled sidebar-section depth1 show">
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/fsdp_qlora.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">FDSP + QLoRA</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/unsloth.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Unsloth</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/torchao.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">PyTorch ao</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/custom_integrations.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Custom Integrations</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/sequence_parallelism.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Sequence Parallelism</span></a>
|
||||
</div>
|
||||
</li>
|
||||
</ul>
|
||||
</li>
|
||||
<li class="sidebar-item sidebar-item-section">
|
||||
<div class="sidebar-item-container">
|
||||
<a class="sidebar-item-text sidebar-link text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-7" role="navigation" aria-expanded="true">
|
||||
<span class="menu-text">Troubleshooting</span></a>
|
||||
<a class="sidebar-item-toggle text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-7" role="navigation" aria-expanded="true" aria-label="Toggle section">
|
||||
<i class="bi bi-chevron-right ms-2"></i>
|
||||
</a>
|
||||
</div>
|
||||
<ul id="quarto-sidebar-section-7" class="collapse list-unstyled sidebar-section depth1 show">
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/faq.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">FAQ</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/debugging.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Debugging</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/nccl.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">NCCL</span></a>
|
||||
</div>
|
||||
</li>
|
||||
</ul>
|
||||
</li>
|
||||
</ul>
|
||||
</div>
|
||||
</nav>
|
||||
<div id="quarto-sidebar-glass" class="quarto-sidebar-collapse-item" data-bs-toggle="collapse" data-bs-target=".quarto-sidebar-collapse-item"></div>
|
||||
<!-- margin-sidebar -->
|
||||
<div id="quarto-margin-sidebar" class="sidebar margin-sidebar">
|
||||
<nav id="TOC" role="doc-toc" class="toc-active">
|
||||
<h2 id="toc-title">On this page</h2>
|
||||
|
||||
<ul>
|
||||
<li><a href="#axolotl.loaders.constants" id="toc-axolotl.loaders.constants" class="nav-link active" data-scroll-target="#axolotl.loaders.constants">loaders.constants</a></li>
|
||||
</ul>
|
||||
</nav>
|
||||
</div>
|
||||
<!-- main -->
|
||||
<main class="content" id="quarto-document-content"><header id="title-block-header" class="quarto-title-block"></header>
|
||||
|
||||
|
||||
|
||||
|
||||
<section id="axolotl.loaders.constants" class="level1">
|
||||
<h1>loaders.constants</h1>
|
||||
<p><code>loaders.constants</code></p>
|
||||
<p>Shared constants for axolotl.loaders module</p>
|
||||
|
||||
|
||||
</section>
|
||||
|
||||
</main> <!-- /main -->
|
||||
<script id="quarto-html-after-body" type="application/javascript">
|
||||
window.document.addEventListener("DOMContentLoaded", function (event) {
|
||||
const icon = "";
|
||||
const anchorJS = new window.AnchorJS();
|
||||
anchorJS.options = {
|
||||
placement: 'right',
|
||||
icon: icon
|
||||
};
|
||||
anchorJS.add('.anchored');
|
||||
const isCodeAnnotation = (el) => {
|
||||
for (const clz of el.classList) {
|
||||
if (clz.startsWith('code-annotation-')) {
|
||||
return true;
|
||||
}
|
||||
}
|
||||
return false;
|
||||
}
|
||||
const onCopySuccess = function(e) {
|
||||
// button target
|
||||
const button = e.trigger;
|
||||
// don't keep focus
|
||||
button.blur();
|
||||
// flash "checked"
|
||||
button.classList.add('code-copy-button-checked');
|
||||
var currentTitle = button.getAttribute("title");
|
||||
button.setAttribute("title", "Copied!");
|
||||
let tooltip;
|
||||
if (window.bootstrap) {
|
||||
button.setAttribute("data-bs-toggle", "tooltip");
|
||||
button.setAttribute("data-bs-placement", "left");
|
||||
button.setAttribute("data-bs-title", "Copied!");
|
||||
tooltip = new bootstrap.Tooltip(button,
|
||||
{ trigger: "manual",
|
||||
customClass: "code-copy-button-tooltip",
|
||||
offset: [0, -8]});
|
||||
tooltip.show();
|
||||
}
|
||||
setTimeout(function() {
|
||||
if (tooltip) {
|
||||
tooltip.hide();
|
||||
button.removeAttribute("data-bs-title");
|
||||
button.removeAttribute("data-bs-toggle");
|
||||
button.removeAttribute("data-bs-placement");
|
||||
}
|
||||
button.setAttribute("title", currentTitle);
|
||||
button.classList.remove('code-copy-button-checked');
|
||||
}, 1000);
|
||||
// clear code selection
|
||||
e.clearSelection();
|
||||
}
|
||||
const getTextToCopy = function(trigger) {
|
||||
const codeEl = trigger.previousElementSibling.cloneNode(true);
|
||||
for (const childEl of codeEl.children) {
|
||||
if (isCodeAnnotation(childEl)) {
|
||||
childEl.remove();
|
||||
}
|
||||
}
|
||||
return codeEl.innerText;
|
||||
}
|
||||
const clipboard = new window.ClipboardJS('.code-copy-button:not([data-in-quarto-modal])', {
|
||||
text: getTextToCopy
|
||||
});
|
||||
clipboard.on('success', onCopySuccess);
|
||||
if (window.document.getElementById('quarto-embedded-source-code-modal')) {
|
||||
const clipboardModal = new window.ClipboardJS('.code-copy-button[data-in-quarto-modal]', {
|
||||
text: getTextToCopy,
|
||||
container: window.document.getElementById('quarto-embedded-source-code-modal')
|
||||
});
|
||||
clipboardModal.on('success', onCopySuccess);
|
||||
}
|
||||
var localhostRegex = new RegExp(/^(?:http|https):\/\/localhost\:?[0-9]*\//);
|
||||
var mailtoRegex = new RegExp(/^mailto:/);
|
||||
var filterRegex = new RegExp("https:\/\/docs\.axolotl\.ai");
|
||||
var isInternal = (href) => {
|
||||
return filterRegex.test(href) || localhostRegex.test(href) || mailtoRegex.test(href);
|
||||
}
|
||||
// Inspect non-navigation links and adorn them if external
|
||||
var links = window.document.querySelectorAll('a[href]:not(.nav-link):not(.navbar-brand):not(.toc-action):not(.sidebar-link):not(.sidebar-item-toggle):not(.pagination-link):not(.no-external):not([aria-hidden]):not(.dropdown-item):not(.quarto-navigation-tool):not(.about-link)');
|
||||
for (var i=0; i<links.length; i++) {
|
||||
const link = links[i];
|
||||
if (!isInternal(link.href)) {
|
||||
// undo the damage that might have been done by quarto-nav.js in the case of
|
||||
// links that we want to consider external
|
||||
if (link.dataset.originalHref !== undefined) {
|
||||
link.href = link.dataset.originalHref;
|
||||
}
|
||||
}
|
||||
}
|
||||
function tippyHover(el, contentFn, onTriggerFn, onUntriggerFn) {
|
||||
const config = {
|
||||
allowHTML: true,
|
||||
maxWidth: 500,
|
||||
delay: 100,
|
||||
arrow: false,
|
||||
appendTo: function(el) {
|
||||
return el.parentElement;
|
||||
},
|
||||
interactive: true,
|
||||
interactiveBorder: 10,
|
||||
theme: 'quarto',
|
||||
placement: 'bottom-start',
|
||||
};
|
||||
if (contentFn) {
|
||||
config.content = contentFn;
|
||||
}
|
||||
if (onTriggerFn) {
|
||||
config.onTrigger = onTriggerFn;
|
||||
}
|
||||
if (onUntriggerFn) {
|
||||
config.onUntrigger = onUntriggerFn;
|
||||
}
|
||||
window.tippy(el, config);
|
||||
}
|
||||
const noterefs = window.document.querySelectorAll('a[role="doc-noteref"]');
|
||||
for (var i=0; i<noterefs.length; i++) {
|
||||
const ref = noterefs[i];
|
||||
tippyHover(ref, function() {
|
||||
// use id or data attribute instead here
|
||||
let href = ref.getAttribute('data-footnote-href') || ref.getAttribute('href');
|
||||
try { href = new URL(href).hash; } catch {}
|
||||
const id = href.replace(/^#\/?/, "");
|
||||
const note = window.document.getElementById(id);
|
||||
if (note) {
|
||||
return note.innerHTML;
|
||||
} else {
|
||||
return "";
|
||||
}
|
||||
});
|
||||
}
|
||||
const xrefs = window.document.querySelectorAll('a.quarto-xref');
|
||||
const processXRef = (id, note) => {
|
||||
// Strip column container classes
|
||||
const stripColumnClz = (el) => {
|
||||
el.classList.remove("page-full", "page-columns");
|
||||
if (el.children) {
|
||||
for (const child of el.children) {
|
||||
stripColumnClz(child);
|
||||
}
|
||||
}
|
||||
}
|
||||
stripColumnClz(note)
|
||||
if (id === null || id.startsWith('sec-')) {
|
||||
// Special case sections, only their first couple elements
|
||||
const container = document.createElement("div");
|
||||
if (note.children && note.children.length > 2) {
|
||||
container.appendChild(note.children[0].cloneNode(true));
|
||||
for (let i = 1; i < note.children.length; i++) {
|
||||
const child = note.children[i];
|
||||
if (child.tagName === "P" && child.innerText === "") {
|
||||
continue;
|
||||
} else {
|
||||
container.appendChild(child.cloneNode(true));
|
||||
break;
|
||||
}
|
||||
}
|
||||
if (window.Quarto?.typesetMath) {
|
||||
window.Quarto.typesetMath(container);
|
||||
}
|
||||
return container.innerHTML
|
||||
} else {
|
||||
if (window.Quarto?.typesetMath) {
|
||||
window.Quarto.typesetMath(note);
|
||||
}
|
||||
return note.innerHTML;
|
||||
}
|
||||
} else {
|
||||
// Remove any anchor links if they are present
|
||||
const anchorLink = note.querySelector('a.anchorjs-link');
|
||||
if (anchorLink) {
|
||||
anchorLink.remove();
|
||||
}
|
||||
if (window.Quarto?.typesetMath) {
|
||||
window.Quarto.typesetMath(note);
|
||||
}
|
||||
if (note.classList.contains("callout")) {
|
||||
return note.outerHTML;
|
||||
} else {
|
||||
return note.innerHTML;
|
||||
}
|
||||
}
|
||||
}
|
||||
for (var i=0; i<xrefs.length; i++) {
|
||||
const xref = xrefs[i];
|
||||
tippyHover(xref, undefined, function(instance) {
|
||||
instance.disable();
|
||||
let url = xref.getAttribute('href');
|
||||
let hash = undefined;
|
||||
if (url.startsWith('#')) {
|
||||
hash = url;
|
||||
} else {
|
||||
try { hash = new URL(url).hash; } catch {}
|
||||
}
|
||||
if (hash) {
|
||||
const id = hash.replace(/^#\/?/, "");
|
||||
const note = window.document.getElementById(id);
|
||||
if (note !== null) {
|
||||
try {
|
||||
const html = processXRef(id, note.cloneNode(true));
|
||||
instance.setContent(html);
|
||||
} finally {
|
||||
instance.enable();
|
||||
instance.show();
|
||||
}
|
||||
} else {
|
||||
// See if we can fetch this
|
||||
fetch(url.split('#')[0])
|
||||
.then(res => res.text())
|
||||
.then(html => {
|
||||
const parser = new DOMParser();
|
||||
const htmlDoc = parser.parseFromString(html, "text/html");
|
||||
const note = htmlDoc.getElementById(id);
|
||||
if (note !== null) {
|
||||
const html = processXRef(id, note);
|
||||
instance.setContent(html);
|
||||
}
|
||||
}).finally(() => {
|
||||
instance.enable();
|
||||
instance.show();
|
||||
});
|
||||
}
|
||||
} else {
|
||||
// See if we can fetch a full url (with no hash to target)
|
||||
// This is a special case and we should probably do some content thinning / targeting
|
||||
fetch(url)
|
||||
.then(res => res.text())
|
||||
.then(html => {
|
||||
const parser = new DOMParser();
|
||||
const htmlDoc = parser.parseFromString(html, "text/html");
|
||||
const note = htmlDoc.querySelector('main.content');
|
||||
if (note !== null) {
|
||||
// This should only happen for chapter cross references
|
||||
// (since there is no id in the URL)
|
||||
// remove the first header
|
||||
if (note.children.length > 0 && note.children[0].tagName === "HEADER") {
|
||||
note.children[0].remove();
|
||||
}
|
||||
const html = processXRef(null, note);
|
||||
instance.setContent(html);
|
||||
}
|
||||
}).finally(() => {
|
||||
instance.enable();
|
||||
instance.show();
|
||||
});
|
||||
}
|
||||
}, function(instance) {
|
||||
});
|
||||
}
|
||||
let selectedAnnoteEl;
|
||||
const selectorForAnnotation = ( cell, annotation) => {
|
||||
let cellAttr = 'data-code-cell="' + cell + '"';
|
||||
let lineAttr = 'data-code-annotation="' + annotation + '"';
|
||||
const selector = 'span[' + cellAttr + '][' + lineAttr + ']';
|
||||
return selector;
|
||||
}
|
||||
const selectCodeLines = (annoteEl) => {
|
||||
const doc = window.document;
|
||||
const targetCell = annoteEl.getAttribute("data-target-cell");
|
||||
const targetAnnotation = annoteEl.getAttribute("data-target-annotation");
|
||||
const annoteSpan = window.document.querySelector(selectorForAnnotation(targetCell, targetAnnotation));
|
||||
const lines = annoteSpan.getAttribute("data-code-lines").split(",");
|
||||
const lineIds = lines.map((line) => {
|
||||
return targetCell + "-" + line;
|
||||
})
|
||||
let top = null;
|
||||
let height = null;
|
||||
let parent = null;
|
||||
if (lineIds.length > 0) {
|
||||
//compute the position of the single el (top and bottom and make a div)
|
||||
const el = window.document.getElementById(lineIds[0]);
|
||||
top = el.offsetTop;
|
||||
height = el.offsetHeight;
|
||||
parent = el.parentElement.parentElement;
|
||||
if (lineIds.length > 1) {
|
||||
const lastEl = window.document.getElementById(lineIds[lineIds.length - 1]);
|
||||
const bottom = lastEl.offsetTop + lastEl.offsetHeight;
|
||||
height = bottom - top;
|
||||
}
|
||||
if (top !== null && height !== null && parent !== null) {
|
||||
// cook up a div (if necessary) and position it
|
||||
let div = window.document.getElementById("code-annotation-line-highlight");
|
||||
if (div === null) {
|
||||
div = window.document.createElement("div");
|
||||
div.setAttribute("id", "code-annotation-line-highlight");
|
||||
div.style.position = 'absolute';
|
||||
parent.appendChild(div);
|
||||
}
|
||||
div.style.top = top - 2 + "px";
|
||||
div.style.height = height + 4 + "px";
|
||||
div.style.left = 0;
|
||||
let gutterDiv = window.document.getElementById("code-annotation-line-highlight-gutter");
|
||||
if (gutterDiv === null) {
|
||||
gutterDiv = window.document.createElement("div");
|
||||
gutterDiv.setAttribute("id", "code-annotation-line-highlight-gutter");
|
||||
gutterDiv.style.position = 'absolute';
|
||||
const codeCell = window.document.getElementById(targetCell);
|
||||
const gutter = codeCell.querySelector('.code-annotation-gutter');
|
||||
gutter.appendChild(gutterDiv);
|
||||
}
|
||||
gutterDiv.style.top = top - 2 + "px";
|
||||
gutterDiv.style.height = height + 4 + "px";
|
||||
}
|
||||
selectedAnnoteEl = annoteEl;
|
||||
}
|
||||
};
|
||||
const unselectCodeLines = () => {
|
||||
const elementsIds = ["code-annotation-line-highlight", "code-annotation-line-highlight-gutter"];
|
||||
elementsIds.forEach((elId) => {
|
||||
const div = window.document.getElementById(elId);
|
||||
if (div) {
|
||||
div.remove();
|
||||
}
|
||||
});
|
||||
selectedAnnoteEl = undefined;
|
||||
};
|
||||
// Handle positioning of the toggle
|
||||
window.addEventListener(
|
||||
"resize",
|
||||
throttle(() => {
|
||||
elRect = undefined;
|
||||
if (selectedAnnoteEl) {
|
||||
selectCodeLines(selectedAnnoteEl);
|
||||
}
|
||||
}, 10)
|
||||
);
|
||||
function throttle(fn, ms) {
|
||||
let throttle = false;
|
||||
let timer;
|
||||
return (...args) => {
|
||||
if(!throttle) { // first call gets through
|
||||
fn.apply(this, args);
|
||||
throttle = true;
|
||||
} else { // all the others get throttled
|
||||
if(timer) clearTimeout(timer); // cancel #2
|
||||
timer = setTimeout(() => {
|
||||
fn.apply(this, args);
|
||||
timer = throttle = false;
|
||||
}, ms);
|
||||
}
|
||||
};
|
||||
}
|
||||
// Attach click handler to the DT
|
||||
const annoteDls = window.document.querySelectorAll('dt[data-target-cell]');
|
||||
for (const annoteDlNode of annoteDls) {
|
||||
annoteDlNode.addEventListener('click', (event) => {
|
||||
const clickedEl = event.target;
|
||||
if (clickedEl !== selectedAnnoteEl) {
|
||||
unselectCodeLines();
|
||||
const activeEl = window.document.querySelector('dt[data-target-cell].code-annotation-active');
|
||||
if (activeEl) {
|
||||
activeEl.classList.remove('code-annotation-active');
|
||||
}
|
||||
selectCodeLines(clickedEl);
|
||||
clickedEl.classList.add('code-annotation-active');
|
||||
} else {
|
||||
// Unselect the line
|
||||
unselectCodeLines();
|
||||
clickedEl.classList.remove('code-annotation-active');
|
||||
}
|
||||
});
|
||||
}
|
||||
const findCites = (el) => {
|
||||
const parentEl = el.parentElement;
|
||||
if (parentEl) {
|
||||
const cites = parentEl.dataset.cites;
|
||||
if (cites) {
|
||||
return {
|
||||
el,
|
||||
cites: cites.split(' ')
|
||||
};
|
||||
} else {
|
||||
return findCites(el.parentElement)
|
||||
}
|
||||
} else {
|
||||
return undefined;
|
||||
}
|
||||
};
|
||||
var bibliorefs = window.document.querySelectorAll('a[role="doc-biblioref"]');
|
||||
for (var i=0; i<bibliorefs.length; i++) {
|
||||
const ref = bibliorefs[i];
|
||||
const citeInfo = findCites(ref);
|
||||
if (citeInfo) {
|
||||
tippyHover(citeInfo.el, function() {
|
||||
var popup = window.document.createElement('div');
|
||||
citeInfo.cites.forEach(function(cite) {
|
||||
var citeDiv = window.document.createElement('div');
|
||||
citeDiv.classList.add('hanging-indent');
|
||||
citeDiv.classList.add('csl-entry');
|
||||
var biblioDiv = window.document.getElementById('ref-' + cite);
|
||||
if (biblioDiv) {
|
||||
citeDiv.innerHTML = biblioDiv.innerHTML;
|
||||
}
|
||||
popup.appendChild(citeDiv);
|
||||
});
|
||||
return popup.innerHTML;
|
||||
});
|
||||
}
|
||||
}
|
||||
});
|
||||
</script>
|
||||
</div> <!-- /content -->
|
||||
|
||||
|
||||
|
||||
|
||||
</body></html>
|
||||
1020
docs/api/loaders.model.html
Normal file
1020
docs/api/loaders.model.html
Normal file
File diff suppressed because it is too large
Load Diff
949
docs/api/loaders.patch_manager.html
Normal file
949
docs/api/loaders.patch_manager.html
Normal file
@@ -0,0 +1,949 @@
|
||||
<!DOCTYPE html>
|
||||
<html xmlns="http://www.w3.org/1999/xhtml" lang="en" xml:lang="en"><head>
|
||||
|
||||
<meta charset="utf-8">
|
||||
<meta name="generator" content="quarto-1.7.31">
|
||||
|
||||
<meta name="viewport" content="width=device-width, initial-scale=1.0, user-scalable=yes">
|
||||
|
||||
|
||||
<title>loaders.patch_manager – Axolotl</title>
|
||||
<style>
|
||||
code{white-space: pre-wrap;}
|
||||
span.smallcaps{font-variant: small-caps;}
|
||||
div.columns{display: flex; gap: min(4vw, 1.5em);}
|
||||
div.column{flex: auto; overflow-x: auto;}
|
||||
div.hanging-indent{margin-left: 1.5em; text-indent: -1.5em;}
|
||||
ul.task-list{list-style: none;}
|
||||
ul.task-list li input[type="checkbox"] {
|
||||
width: 0.8em;
|
||||
margin: 0 0.8em 0.2em -1em; /* quarto-specific, see https://github.com/quarto-dev/quarto-cli/issues/4556 */
|
||||
vertical-align: middle;
|
||||
}
|
||||
/* CSS for syntax highlighting */
|
||||
html { -webkit-text-size-adjust: 100%; }
|
||||
pre > code.sourceCode { white-space: pre; position: relative; }
|
||||
pre > code.sourceCode > span { display: inline-block; line-height: 1.25; }
|
||||
pre > code.sourceCode > span:empty { height: 1.2em; }
|
||||
.sourceCode { overflow: visible; }
|
||||
code.sourceCode > span { color: inherit; text-decoration: inherit; }
|
||||
div.sourceCode { margin: 1em 0; }
|
||||
pre.sourceCode { margin: 0; }
|
||||
@media screen {
|
||||
div.sourceCode { overflow: auto; }
|
||||
}
|
||||
@media print {
|
||||
pre > code.sourceCode { white-space: pre-wrap; }
|
||||
pre > code.sourceCode > span { text-indent: -5em; padding-left: 5em; }
|
||||
}
|
||||
pre.numberSource code
|
||||
{ counter-reset: source-line 0; }
|
||||
pre.numberSource code > span
|
||||
{ position: relative; left: -4em; counter-increment: source-line; }
|
||||
pre.numberSource code > span > a:first-child::before
|
||||
{ content: counter(source-line);
|
||||
position: relative; left: -1em; text-align: right; vertical-align: baseline;
|
||||
border: none; display: inline-block;
|
||||
-webkit-touch-callout: none; -webkit-user-select: none;
|
||||
-khtml-user-select: none; -moz-user-select: none;
|
||||
-ms-user-select: none; user-select: none;
|
||||
padding: 0 4px; width: 4em;
|
||||
}
|
||||
pre.numberSource { margin-left: 3em; padding-left: 4px; }
|
||||
div.sourceCode
|
||||
{ }
|
||||
@media screen {
|
||||
pre > code.sourceCode > span > a:first-child::before { text-decoration: underline; }
|
||||
}
|
||||
</style>
|
||||
|
||||
|
||||
<script src="../../site_libs/quarto-nav/quarto-nav.js"></script>
|
||||
<script src="../../site_libs/clipboard/clipboard.min.js"></script>
|
||||
<script src="../../site_libs/quarto-search/autocomplete.umd.js"></script>
|
||||
<script src="../../site_libs/quarto-search/fuse.min.js"></script>
|
||||
<script src="../../site_libs/quarto-search/quarto-search.js"></script>
|
||||
<meta name="quarto:offset" content="../../">
|
||||
<link href="../../favicon.jpg" rel="icon" type="image/jpeg">
|
||||
<script src="../../site_libs/quarto-html/quarto.js" type="module"></script>
|
||||
<script src="../../site_libs/quarto-html/tabsets/tabsets.js" type="module"></script>
|
||||
<script src="../../site_libs/quarto-html/popper.min.js"></script>
|
||||
<script src="../../site_libs/quarto-html/tippy.umd.min.js"></script>
|
||||
<script src="../../site_libs/quarto-html/anchor.min.js"></script>
|
||||
<link href="../../site_libs/quarto-html/tippy.css" rel="stylesheet">
|
||||
<link href="../../site_libs/quarto-html/quarto-syntax-highlighting-dark-8ef56b68f8fa1e9d2ba328e99e439f80.css" rel="stylesheet" id="quarto-text-highlighting-styles">
|
||||
<script src="../../site_libs/bootstrap/bootstrap.min.js"></script>
|
||||
<link href="../../site_libs/bootstrap/bootstrap-icons.css" rel="stylesheet">
|
||||
<link href="../../site_libs/bootstrap/bootstrap-2288ecdcbf81d2ab6432743cedd71d9a.min.css" rel="stylesheet" append-hash="true" id="quarto-bootstrap" data-mode="dark">
|
||||
<script id="quarto-search-options" type="application/json">{
|
||||
"location": "navbar",
|
||||
"copy-button": false,
|
||||
"collapse-after": 3,
|
||||
"panel-placement": "end",
|
||||
"type": "overlay",
|
||||
"limit": 50,
|
||||
"keyboard-shortcut": [
|
||||
"f",
|
||||
"/",
|
||||
"s"
|
||||
],
|
||||
"show-item-context": false,
|
||||
"language": {
|
||||
"search-no-results-text": "No results",
|
||||
"search-matching-documents-text": "matching documents",
|
||||
"search-copy-link-title": "Copy link to search",
|
||||
"search-hide-matches-text": "Hide additional matches",
|
||||
"search-more-match-text": "more match in this document",
|
||||
"search-more-matches-text": "more matches in this document",
|
||||
"search-clear-button-title": "Clear",
|
||||
"search-text-placeholder": "",
|
||||
"search-detached-cancel-button-title": "Cancel",
|
||||
"search-submit-button-title": "Submit",
|
||||
"search-label": "Search"
|
||||
}
|
||||
}</script>
|
||||
|
||||
|
||||
<link rel="stylesheet" href="../../styles.css">
|
||||
</head>
|
||||
|
||||
<body class="nav-sidebar docked nav-fixed quarto-light">
|
||||
|
||||
<div id="quarto-search-results"></div>
|
||||
<header id="quarto-header" class="headroom fixed-top">
|
||||
<nav class="navbar navbar-expand " data-bs-theme="dark">
|
||||
<div class="navbar-container container-fluid">
|
||||
<div class="navbar-brand-container mx-auto">
|
||||
<a href="../../index.html" class="navbar-brand navbar-brand-logo">
|
||||
<img src="../../image/axolotl_logo_digital_white.svg" alt="" class="navbar-logo">
|
||||
</a>
|
||||
</div>
|
||||
<div class="quarto-navbar-tools tools-wide tools-end">
|
||||
<a href="https://twitter.com/axolotl_ai" title="" class="quarto-navigation-tool px-1" aria-label=""><i class="bi bi-twitter"></i></a>
|
||||
<a href="https://github.com/axolotl-ai-cloud/axolotl/" title="" class="quarto-navigation-tool px-1" aria-label=""><i class="bi bi-github"></i></a>
|
||||
<a href="https://discord.gg/7m9sfhzaf3" title="" class="quarto-navigation-tool px-1" aria-label=""><i class="bi bi-discord"></i></a>
|
||||
</div>
|
||||
<div id="quarto-search" class="" title="Search"></div>
|
||||
</div> <!-- /container-fluid -->
|
||||
</nav>
|
||||
<nav class="quarto-secondary-nav">
|
||||
<div class="container-fluid d-flex">
|
||||
<button type="button" class="quarto-btn-toggle btn" data-bs-toggle="collapse" role="button" data-bs-target=".quarto-sidebar-collapse-item" aria-controls="quarto-sidebar" aria-expanded="false" aria-label="Toggle sidebar navigation" onclick="if (window.quartoToggleHeadroom) { window.quartoToggleHeadroom(); }">
|
||||
<i class="bi bi-layout-text-sidebar-reverse"></i>
|
||||
</button>
|
||||
<nav class="quarto-page-breadcrumbs" aria-label="breadcrumb"><ol class="breadcrumb"></ol></nav>
|
||||
<a class="flex-grow-1" role="navigation" data-bs-toggle="collapse" data-bs-target=".quarto-sidebar-collapse-item" aria-controls="quarto-sidebar" aria-expanded="false" aria-label="Toggle sidebar navigation" onclick="if (window.quartoToggleHeadroom) { window.quartoToggleHeadroom(); }">
|
||||
</a>
|
||||
</div>
|
||||
</nav>
|
||||
</header>
|
||||
<!-- content -->
|
||||
<div id="quarto-content" class="quarto-container page-columns page-rows-contents page-layout-article page-navbar">
|
||||
<!-- sidebar -->
|
||||
<nav id="quarto-sidebar" class="sidebar collapse collapse-horizontal quarto-sidebar-collapse-item sidebar-navigation docked overflow-auto">
|
||||
<div class="sidebar-menu-container">
|
||||
<ul class="list-unstyled mt-1">
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../index.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Home</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item sidebar-item-section">
|
||||
<div class="sidebar-item-container">
|
||||
<a class="sidebar-item-text sidebar-link text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-1" role="navigation" aria-expanded="true">
|
||||
<span class="menu-text">Getting Started</span></a>
|
||||
<a class="sidebar-item-toggle text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-1" role="navigation" aria-expanded="true" aria-label="Toggle section">
|
||||
<i class="bi bi-chevron-right ms-2"></i>
|
||||
</a>
|
||||
</div>
|
||||
<ul id="quarto-sidebar-section-1" class="collapse list-unstyled sidebar-section depth1 show">
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/getting-started.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Quickstart</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/installation.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Installation</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/inference.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Inference and Merging</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/cli.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Command Line Interface (CLI)</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/config.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Config Reference</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/api" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">API Reference</span></a>
|
||||
</div>
|
||||
</li>
|
||||
</ul>
|
||||
</li>
|
||||
<li class="sidebar-item sidebar-item-section">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/dataset-formats/index.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Dataset Formats</span></a>
|
||||
<a class="sidebar-item-toggle text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-2" role="navigation" aria-expanded="true" aria-label="Toggle section">
|
||||
<i class="bi bi-chevron-right ms-2"></i>
|
||||
</a>
|
||||
</div>
|
||||
<ul id="quarto-sidebar-section-2" class="collapse list-unstyled sidebar-section depth1 show">
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/dataset-formats/pretraining.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Pre-training</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/dataset-formats/inst_tune.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Instruction Tuning</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/dataset-formats/conversation.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Conversation</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/dataset-formats/stepwise_supervised.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Stepwise Supervised Format</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/dataset-formats/template_free.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Template-Free</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/dataset-formats/tokenized.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Custom Pre-Tokenized Dataset</span></a>
|
||||
</div>
|
||||
</li>
|
||||
</ul>
|
||||
</li>
|
||||
<li class="sidebar-item sidebar-item-section">
|
||||
<div class="sidebar-item-container">
|
||||
<a class="sidebar-item-text sidebar-link text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-3" role="navigation" aria-expanded="true">
|
||||
<span class="menu-text">Deployments</span></a>
|
||||
<a class="sidebar-item-toggle text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-3" role="navigation" aria-expanded="true" aria-label="Toggle section">
|
||||
<i class="bi bi-chevron-right ms-2"></i>
|
||||
</a>
|
||||
</div>
|
||||
<ul id="quarto-sidebar-section-3" class="collapse list-unstyled sidebar-section depth1 show">
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/docker.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Docker</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/multi-gpu.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Multi-GPU</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/multi-node.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Multi Node</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/ray-integration.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Ray Train</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/amd_hpc.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">AMD GPUs on HPC Systems</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/mac.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Mac M-series</span></a>
|
||||
</div>
|
||||
</li>
|
||||
</ul>
|
||||
</li>
|
||||
<li class="sidebar-item sidebar-item-section">
|
||||
<div class="sidebar-item-container">
|
||||
<a class="sidebar-item-text sidebar-link text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-4" role="navigation" aria-expanded="true">
|
||||
<span class="menu-text">How To Guides</span></a>
|
||||
<a class="sidebar-item-toggle text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-4" role="navigation" aria-expanded="true" aria-label="Toggle section">
|
||||
<i class="bi bi-chevron-right ms-2"></i>
|
||||
</a>
|
||||
</div>
|
||||
<ul id="quarto-sidebar-section-4" class="collapse list-unstyled sidebar-section depth1 show">
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/multimodal.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">MultiModal / Vision Language Models (BETA)</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/rlhf.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">RLHF (Beta)</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/reward_modelling.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Reward Modelling</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/lr_groups.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Learning Rate Groups</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/lora_optims.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">LoRA Optimizations</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/dataset_loading.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Dataset Loading</span></a>
|
||||
</div>
|
||||
</li>
|
||||
</ul>
|
||||
</li>
|
||||
<li class="sidebar-item sidebar-item-section">
|
||||
<div class="sidebar-item-container">
|
||||
<a class="sidebar-item-text sidebar-link text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-5" role="navigation" aria-expanded="true">
|
||||
<span class="menu-text">Core Concepts</span></a>
|
||||
<a class="sidebar-item-toggle text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-5" role="navigation" aria-expanded="true" aria-label="Toggle section">
|
||||
<i class="bi bi-chevron-right ms-2"></i>
|
||||
</a>
|
||||
</div>
|
||||
<ul id="quarto-sidebar-section-5" class="collapse list-unstyled sidebar-section depth1 show">
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/batch_vs_grad.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Batch size vs Gradient accumulation</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/dataset_preprocessing.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Dataset Preprocessing</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/multipack.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Multipack (Sample Packing)</span></a>
|
||||
</div>
|
||||
</li>
|
||||
</ul>
|
||||
</li>
|
||||
<li class="sidebar-item sidebar-item-section">
|
||||
<div class="sidebar-item-container">
|
||||
<a class="sidebar-item-text sidebar-link text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-6" role="navigation" aria-expanded="true">
|
||||
<span class="menu-text">Advanced Features</span></a>
|
||||
<a class="sidebar-item-toggle text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-6" role="navigation" aria-expanded="true" aria-label="Toggle section">
|
||||
<i class="bi bi-chevron-right ms-2"></i>
|
||||
</a>
|
||||
</div>
|
||||
<ul id="quarto-sidebar-section-6" class="collapse list-unstyled sidebar-section depth1 show">
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/fsdp_qlora.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">FDSP + QLoRA</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/unsloth.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Unsloth</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/torchao.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">PyTorch ao</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/custom_integrations.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Custom Integrations</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/sequence_parallelism.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Sequence Parallelism</span></a>
|
||||
</div>
|
||||
</li>
|
||||
</ul>
|
||||
</li>
|
||||
<li class="sidebar-item sidebar-item-section">
|
||||
<div class="sidebar-item-container">
|
||||
<a class="sidebar-item-text sidebar-link text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-7" role="navigation" aria-expanded="true">
|
||||
<span class="menu-text">Troubleshooting</span></a>
|
||||
<a class="sidebar-item-toggle text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-7" role="navigation" aria-expanded="true" aria-label="Toggle section">
|
||||
<i class="bi bi-chevron-right ms-2"></i>
|
||||
</a>
|
||||
</div>
|
||||
<ul id="quarto-sidebar-section-7" class="collapse list-unstyled sidebar-section depth1 show">
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/faq.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">FAQ</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/debugging.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Debugging</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/nccl.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">NCCL</span></a>
|
||||
</div>
|
||||
</li>
|
||||
</ul>
|
||||
</li>
|
||||
</ul>
|
||||
</div>
|
||||
</nav>
|
||||
<div id="quarto-sidebar-glass" class="quarto-sidebar-collapse-item" data-bs-toggle="collapse" data-bs-target=".quarto-sidebar-collapse-item"></div>
|
||||
<!-- margin-sidebar -->
|
||||
<div id="quarto-margin-sidebar" class="sidebar margin-sidebar">
|
||||
<nav id="TOC" role="doc-toc" class="toc-active">
|
||||
<h2 id="toc-title">On this page</h2>
|
||||
|
||||
<ul>
|
||||
<li><a href="#axolotl.loaders.patch_manager" id="toc-axolotl.loaders.patch_manager" class="nav-link active" data-scroll-target="#axolotl.loaders.patch_manager">loaders.patch_manager</a>
|
||||
<ul class="collapse">
|
||||
<li><a href="#classes" id="toc-classes" class="nav-link" data-scroll-target="#classes">Classes</a>
|
||||
<ul class="collapse">
|
||||
<li><a href="#axolotl.loaders.patch_manager.PatchManager" id="toc-axolotl.loaders.patch_manager.PatchManager" class="nav-link" data-scroll-target="#axolotl.loaders.patch_manager.PatchManager">PatchManager</a></li>
|
||||
</ul></li>
|
||||
</ul></li>
|
||||
</ul>
|
||||
</nav>
|
||||
</div>
|
||||
<!-- main -->
|
||||
<main class="content" id="quarto-document-content"><header id="title-block-header" class="quarto-title-block"></header>
|
||||
|
||||
|
||||
|
||||
|
||||
<section id="axolotl.loaders.patch_manager" class="level1">
|
||||
<h1>loaders.patch_manager</h1>
|
||||
<p><code>loaders.patch_manager</code></p>
|
||||
<p>Patch manager class implementation to complement <code>axolotl.loaders.ModelLoader</code>.</p>
|
||||
<p>Applies pre- and post-model load patches for various fixes and optimizations.</p>
|
||||
<section id="classes" class="level2">
|
||||
<h2 class="anchored" data-anchor-id="classes">Classes</h2>
|
||||
<table class="caption-top table">
|
||||
<thead>
|
||||
<tr class="header">
|
||||
<th>Name</th>
|
||||
<th>Description</th>
|
||||
</tr>
|
||||
</thead>
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td><a href="#axolotl.loaders.patch_manager.PatchManager">PatchManager</a></td>
|
||||
<td>Manages the application of patches during the model loading process.</td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
<section id="axolotl.loaders.patch_manager.PatchManager" class="level3">
|
||||
<h3 class="anchored" data-anchor-id="axolotl.loaders.patch_manager.PatchManager">PatchManager</h3>
|
||||
<div class="sourceCode" id="cb1"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb1-1"><a href="#cb1-1" aria-hidden="true" tabindex="-1"></a>loaders.patch_manager.PatchManager(<span class="va">self</span>, cfg, model_config, inference<span class="op">=</span><span class="va">False</span>)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Manages the application of patches during the model loading process.</p>
|
||||
<section id="attributes" class="level4">
|
||||
<h4 class="anchored" data-anchor-id="attributes">Attributes</h4>
|
||||
<table class="caption-top table">
|
||||
<thead>
|
||||
<tr class="header">
|
||||
<th>Name</th>
|
||||
<th>Description</th>
|
||||
</tr>
|
||||
</thead>
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td><a href="#axolotl.loaders.patch_manager.PatchManager.has_flash_attn">has_flash_attn</a></td>
|
||||
<td>Check if flash attention is installed.</td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
</section>
|
||||
<section id="methods" class="level4">
|
||||
<h4 class="anchored" data-anchor-id="methods">Methods</h4>
|
||||
<table class="caption-top table">
|
||||
<thead>
|
||||
<tr class="header">
|
||||
<th>Name</th>
|
||||
<th>Description</th>
|
||||
</tr>
|
||||
</thead>
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td><a href="#axolotl.loaders.patch_manager.PatchManager.apply_post_model_load_patches">apply_post_model_load_patches</a></td>
|
||||
<td>Apply patches that require the model instance.</td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td><a href="#axolotl.loaders.patch_manager.PatchManager.apply_pre_model_load_patches">apply_pre_model_load_patches</a></td>
|
||||
<td>Apply pre-model load patches based on config.</td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
<section id="axolotl.loaders.patch_manager.PatchManager.apply_post_model_load_patches" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.loaders.patch_manager.PatchManager.apply_post_model_load_patches">apply_post_model_load_patches</h5>
|
||||
<div class="sourceCode" id="cb2"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb2-1"><a href="#cb2-1" aria-hidden="true" tabindex="-1"></a>loaders.patch_manager.PatchManager.apply_post_model_load_patches(model)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Apply patches that require the model instance.</p>
|
||||
</section>
|
||||
<section id="axolotl.loaders.patch_manager.PatchManager.apply_pre_model_load_patches" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.loaders.patch_manager.PatchManager.apply_pre_model_load_patches">apply_pre_model_load_patches</h5>
|
||||
<div class="sourceCode" id="cb3"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb3-1"><a href="#cb3-1" aria-hidden="true" tabindex="-1"></a>loaders.patch_manager.PatchManager.apply_pre_model_load_patches()</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Apply pre-model load patches based on config.</p>
|
||||
|
||||
|
||||
</section>
|
||||
</section>
|
||||
</section>
|
||||
</section>
|
||||
</section>
|
||||
|
||||
</main> <!-- /main -->
|
||||
<script id="quarto-html-after-body" type="application/javascript">
|
||||
window.document.addEventListener("DOMContentLoaded", function (event) {
|
||||
const icon = "";
|
||||
const anchorJS = new window.AnchorJS();
|
||||
anchorJS.options = {
|
||||
placement: 'right',
|
||||
icon: icon
|
||||
};
|
||||
anchorJS.add('.anchored');
|
||||
const isCodeAnnotation = (el) => {
|
||||
for (const clz of el.classList) {
|
||||
if (clz.startsWith('code-annotation-')) {
|
||||
return true;
|
||||
}
|
||||
}
|
||||
return false;
|
||||
}
|
||||
const onCopySuccess = function(e) {
|
||||
// button target
|
||||
const button = e.trigger;
|
||||
// don't keep focus
|
||||
button.blur();
|
||||
// flash "checked"
|
||||
button.classList.add('code-copy-button-checked');
|
||||
var currentTitle = button.getAttribute("title");
|
||||
button.setAttribute("title", "Copied!");
|
||||
let tooltip;
|
||||
if (window.bootstrap) {
|
||||
button.setAttribute("data-bs-toggle", "tooltip");
|
||||
button.setAttribute("data-bs-placement", "left");
|
||||
button.setAttribute("data-bs-title", "Copied!");
|
||||
tooltip = new bootstrap.Tooltip(button,
|
||||
{ trigger: "manual",
|
||||
customClass: "code-copy-button-tooltip",
|
||||
offset: [0, -8]});
|
||||
tooltip.show();
|
||||
}
|
||||
setTimeout(function() {
|
||||
if (tooltip) {
|
||||
tooltip.hide();
|
||||
button.removeAttribute("data-bs-title");
|
||||
button.removeAttribute("data-bs-toggle");
|
||||
button.removeAttribute("data-bs-placement");
|
||||
}
|
||||
button.setAttribute("title", currentTitle);
|
||||
button.classList.remove('code-copy-button-checked');
|
||||
}, 1000);
|
||||
// clear code selection
|
||||
e.clearSelection();
|
||||
}
|
||||
const getTextToCopy = function(trigger) {
|
||||
const codeEl = trigger.previousElementSibling.cloneNode(true);
|
||||
for (const childEl of codeEl.children) {
|
||||
if (isCodeAnnotation(childEl)) {
|
||||
childEl.remove();
|
||||
}
|
||||
}
|
||||
return codeEl.innerText;
|
||||
}
|
||||
const clipboard = new window.ClipboardJS('.code-copy-button:not([data-in-quarto-modal])', {
|
||||
text: getTextToCopy
|
||||
});
|
||||
clipboard.on('success', onCopySuccess);
|
||||
if (window.document.getElementById('quarto-embedded-source-code-modal')) {
|
||||
const clipboardModal = new window.ClipboardJS('.code-copy-button[data-in-quarto-modal]', {
|
||||
text: getTextToCopy,
|
||||
container: window.document.getElementById('quarto-embedded-source-code-modal')
|
||||
});
|
||||
clipboardModal.on('success', onCopySuccess);
|
||||
}
|
||||
var localhostRegex = new RegExp(/^(?:http|https):\/\/localhost\:?[0-9]*\//);
|
||||
var mailtoRegex = new RegExp(/^mailto:/);
|
||||
var filterRegex = new RegExp("https:\/\/docs\.axolotl\.ai");
|
||||
var isInternal = (href) => {
|
||||
return filterRegex.test(href) || localhostRegex.test(href) || mailtoRegex.test(href);
|
||||
}
|
||||
// Inspect non-navigation links and adorn them if external
|
||||
var links = window.document.querySelectorAll('a[href]:not(.nav-link):not(.navbar-brand):not(.toc-action):not(.sidebar-link):not(.sidebar-item-toggle):not(.pagination-link):not(.no-external):not([aria-hidden]):not(.dropdown-item):not(.quarto-navigation-tool):not(.about-link)');
|
||||
for (var i=0; i<links.length; i++) {
|
||||
const link = links[i];
|
||||
if (!isInternal(link.href)) {
|
||||
// undo the damage that might have been done by quarto-nav.js in the case of
|
||||
// links that we want to consider external
|
||||
if (link.dataset.originalHref !== undefined) {
|
||||
link.href = link.dataset.originalHref;
|
||||
}
|
||||
}
|
||||
}
|
||||
function tippyHover(el, contentFn, onTriggerFn, onUntriggerFn) {
|
||||
const config = {
|
||||
allowHTML: true,
|
||||
maxWidth: 500,
|
||||
delay: 100,
|
||||
arrow: false,
|
||||
appendTo: function(el) {
|
||||
return el.parentElement;
|
||||
},
|
||||
interactive: true,
|
||||
interactiveBorder: 10,
|
||||
theme: 'quarto',
|
||||
placement: 'bottom-start',
|
||||
};
|
||||
if (contentFn) {
|
||||
config.content = contentFn;
|
||||
}
|
||||
if (onTriggerFn) {
|
||||
config.onTrigger = onTriggerFn;
|
||||
}
|
||||
if (onUntriggerFn) {
|
||||
config.onUntrigger = onUntriggerFn;
|
||||
}
|
||||
window.tippy(el, config);
|
||||
}
|
||||
const noterefs = window.document.querySelectorAll('a[role="doc-noteref"]');
|
||||
for (var i=0; i<noterefs.length; i++) {
|
||||
const ref = noterefs[i];
|
||||
tippyHover(ref, function() {
|
||||
// use id or data attribute instead here
|
||||
let href = ref.getAttribute('data-footnote-href') || ref.getAttribute('href');
|
||||
try { href = new URL(href).hash; } catch {}
|
||||
const id = href.replace(/^#\/?/, "");
|
||||
const note = window.document.getElementById(id);
|
||||
if (note) {
|
||||
return note.innerHTML;
|
||||
} else {
|
||||
return "";
|
||||
}
|
||||
});
|
||||
}
|
||||
const xrefs = window.document.querySelectorAll('a.quarto-xref');
|
||||
const processXRef = (id, note) => {
|
||||
// Strip column container classes
|
||||
const stripColumnClz = (el) => {
|
||||
el.classList.remove("page-full", "page-columns");
|
||||
if (el.children) {
|
||||
for (const child of el.children) {
|
||||
stripColumnClz(child);
|
||||
}
|
||||
}
|
||||
}
|
||||
stripColumnClz(note)
|
||||
if (id === null || id.startsWith('sec-')) {
|
||||
// Special case sections, only their first couple elements
|
||||
const container = document.createElement("div");
|
||||
if (note.children && note.children.length > 2) {
|
||||
container.appendChild(note.children[0].cloneNode(true));
|
||||
for (let i = 1; i < note.children.length; i++) {
|
||||
const child = note.children[i];
|
||||
if (child.tagName === "P" && child.innerText === "") {
|
||||
continue;
|
||||
} else {
|
||||
container.appendChild(child.cloneNode(true));
|
||||
break;
|
||||
}
|
||||
}
|
||||
if (window.Quarto?.typesetMath) {
|
||||
window.Quarto.typesetMath(container);
|
||||
}
|
||||
return container.innerHTML
|
||||
} else {
|
||||
if (window.Quarto?.typesetMath) {
|
||||
window.Quarto.typesetMath(note);
|
||||
}
|
||||
return note.innerHTML;
|
||||
}
|
||||
} else {
|
||||
// Remove any anchor links if they are present
|
||||
const anchorLink = note.querySelector('a.anchorjs-link');
|
||||
if (anchorLink) {
|
||||
anchorLink.remove();
|
||||
}
|
||||
if (window.Quarto?.typesetMath) {
|
||||
window.Quarto.typesetMath(note);
|
||||
}
|
||||
if (note.classList.contains("callout")) {
|
||||
return note.outerHTML;
|
||||
} else {
|
||||
return note.innerHTML;
|
||||
}
|
||||
}
|
||||
}
|
||||
for (var i=0; i<xrefs.length; i++) {
|
||||
const xref = xrefs[i];
|
||||
tippyHover(xref, undefined, function(instance) {
|
||||
instance.disable();
|
||||
let url = xref.getAttribute('href');
|
||||
let hash = undefined;
|
||||
if (url.startsWith('#')) {
|
||||
hash = url;
|
||||
} else {
|
||||
try { hash = new URL(url).hash; } catch {}
|
||||
}
|
||||
if (hash) {
|
||||
const id = hash.replace(/^#\/?/, "");
|
||||
const note = window.document.getElementById(id);
|
||||
if (note !== null) {
|
||||
try {
|
||||
const html = processXRef(id, note.cloneNode(true));
|
||||
instance.setContent(html);
|
||||
} finally {
|
||||
instance.enable();
|
||||
instance.show();
|
||||
}
|
||||
} else {
|
||||
// See if we can fetch this
|
||||
fetch(url.split('#')[0])
|
||||
.then(res => res.text())
|
||||
.then(html => {
|
||||
const parser = new DOMParser();
|
||||
const htmlDoc = parser.parseFromString(html, "text/html");
|
||||
const note = htmlDoc.getElementById(id);
|
||||
if (note !== null) {
|
||||
const html = processXRef(id, note);
|
||||
instance.setContent(html);
|
||||
}
|
||||
}).finally(() => {
|
||||
instance.enable();
|
||||
instance.show();
|
||||
});
|
||||
}
|
||||
} else {
|
||||
// See if we can fetch a full url (with no hash to target)
|
||||
// This is a special case and we should probably do some content thinning / targeting
|
||||
fetch(url)
|
||||
.then(res => res.text())
|
||||
.then(html => {
|
||||
const parser = new DOMParser();
|
||||
const htmlDoc = parser.parseFromString(html, "text/html");
|
||||
const note = htmlDoc.querySelector('main.content');
|
||||
if (note !== null) {
|
||||
// This should only happen for chapter cross references
|
||||
// (since there is no id in the URL)
|
||||
// remove the first header
|
||||
if (note.children.length > 0 && note.children[0].tagName === "HEADER") {
|
||||
note.children[0].remove();
|
||||
}
|
||||
const html = processXRef(null, note);
|
||||
instance.setContent(html);
|
||||
}
|
||||
}).finally(() => {
|
||||
instance.enable();
|
||||
instance.show();
|
||||
});
|
||||
}
|
||||
}, function(instance) {
|
||||
});
|
||||
}
|
||||
let selectedAnnoteEl;
|
||||
const selectorForAnnotation = ( cell, annotation) => {
|
||||
let cellAttr = 'data-code-cell="' + cell + '"';
|
||||
let lineAttr = 'data-code-annotation="' + annotation + '"';
|
||||
const selector = 'span[' + cellAttr + '][' + lineAttr + ']';
|
||||
return selector;
|
||||
}
|
||||
const selectCodeLines = (annoteEl) => {
|
||||
const doc = window.document;
|
||||
const targetCell = annoteEl.getAttribute("data-target-cell");
|
||||
const targetAnnotation = annoteEl.getAttribute("data-target-annotation");
|
||||
const annoteSpan = window.document.querySelector(selectorForAnnotation(targetCell, targetAnnotation));
|
||||
const lines = annoteSpan.getAttribute("data-code-lines").split(",");
|
||||
const lineIds = lines.map((line) => {
|
||||
return targetCell + "-" + line;
|
||||
})
|
||||
let top = null;
|
||||
let height = null;
|
||||
let parent = null;
|
||||
if (lineIds.length > 0) {
|
||||
//compute the position of the single el (top and bottom and make a div)
|
||||
const el = window.document.getElementById(lineIds[0]);
|
||||
top = el.offsetTop;
|
||||
height = el.offsetHeight;
|
||||
parent = el.parentElement.parentElement;
|
||||
if (lineIds.length > 1) {
|
||||
const lastEl = window.document.getElementById(lineIds[lineIds.length - 1]);
|
||||
const bottom = lastEl.offsetTop + lastEl.offsetHeight;
|
||||
height = bottom - top;
|
||||
}
|
||||
if (top !== null && height !== null && parent !== null) {
|
||||
// cook up a div (if necessary) and position it
|
||||
let div = window.document.getElementById("code-annotation-line-highlight");
|
||||
if (div === null) {
|
||||
div = window.document.createElement("div");
|
||||
div.setAttribute("id", "code-annotation-line-highlight");
|
||||
div.style.position = 'absolute';
|
||||
parent.appendChild(div);
|
||||
}
|
||||
div.style.top = top - 2 + "px";
|
||||
div.style.height = height + 4 + "px";
|
||||
div.style.left = 0;
|
||||
let gutterDiv = window.document.getElementById("code-annotation-line-highlight-gutter");
|
||||
if (gutterDiv === null) {
|
||||
gutterDiv = window.document.createElement("div");
|
||||
gutterDiv.setAttribute("id", "code-annotation-line-highlight-gutter");
|
||||
gutterDiv.style.position = 'absolute';
|
||||
const codeCell = window.document.getElementById(targetCell);
|
||||
const gutter = codeCell.querySelector('.code-annotation-gutter');
|
||||
gutter.appendChild(gutterDiv);
|
||||
}
|
||||
gutterDiv.style.top = top - 2 + "px";
|
||||
gutterDiv.style.height = height + 4 + "px";
|
||||
}
|
||||
selectedAnnoteEl = annoteEl;
|
||||
}
|
||||
};
|
||||
const unselectCodeLines = () => {
|
||||
const elementsIds = ["code-annotation-line-highlight", "code-annotation-line-highlight-gutter"];
|
||||
elementsIds.forEach((elId) => {
|
||||
const div = window.document.getElementById(elId);
|
||||
if (div) {
|
||||
div.remove();
|
||||
}
|
||||
});
|
||||
selectedAnnoteEl = undefined;
|
||||
};
|
||||
// Handle positioning of the toggle
|
||||
window.addEventListener(
|
||||
"resize",
|
||||
throttle(() => {
|
||||
elRect = undefined;
|
||||
if (selectedAnnoteEl) {
|
||||
selectCodeLines(selectedAnnoteEl);
|
||||
}
|
||||
}, 10)
|
||||
);
|
||||
function throttle(fn, ms) {
|
||||
let throttle = false;
|
||||
let timer;
|
||||
return (...args) => {
|
||||
if(!throttle) { // first call gets through
|
||||
fn.apply(this, args);
|
||||
throttle = true;
|
||||
} else { // all the others get throttled
|
||||
if(timer) clearTimeout(timer); // cancel #2
|
||||
timer = setTimeout(() => {
|
||||
fn.apply(this, args);
|
||||
timer = throttle = false;
|
||||
}, ms);
|
||||
}
|
||||
};
|
||||
}
|
||||
// Attach click handler to the DT
|
||||
const annoteDls = window.document.querySelectorAll('dt[data-target-cell]');
|
||||
for (const annoteDlNode of annoteDls) {
|
||||
annoteDlNode.addEventListener('click', (event) => {
|
||||
const clickedEl = event.target;
|
||||
if (clickedEl !== selectedAnnoteEl) {
|
||||
unselectCodeLines();
|
||||
const activeEl = window.document.querySelector('dt[data-target-cell].code-annotation-active');
|
||||
if (activeEl) {
|
||||
activeEl.classList.remove('code-annotation-active');
|
||||
}
|
||||
selectCodeLines(clickedEl);
|
||||
clickedEl.classList.add('code-annotation-active');
|
||||
} else {
|
||||
// Unselect the line
|
||||
unselectCodeLines();
|
||||
clickedEl.classList.remove('code-annotation-active');
|
||||
}
|
||||
});
|
||||
}
|
||||
const findCites = (el) => {
|
||||
const parentEl = el.parentElement;
|
||||
if (parentEl) {
|
||||
const cites = parentEl.dataset.cites;
|
||||
if (cites) {
|
||||
return {
|
||||
el,
|
||||
cites: cites.split(' ')
|
||||
};
|
||||
} else {
|
||||
return findCites(el.parentElement)
|
||||
}
|
||||
} else {
|
||||
return undefined;
|
||||
}
|
||||
};
|
||||
var bibliorefs = window.document.querySelectorAll('a[role="doc-biblioref"]');
|
||||
for (var i=0; i<bibliorefs.length; i++) {
|
||||
const ref = bibliorefs[i];
|
||||
const citeInfo = findCites(ref);
|
||||
if (citeInfo) {
|
||||
tippyHover(citeInfo.el, function() {
|
||||
var popup = window.document.createElement('div');
|
||||
citeInfo.cites.forEach(function(cite) {
|
||||
var citeDiv = window.document.createElement('div');
|
||||
citeDiv.classList.add('hanging-indent');
|
||||
citeDiv.classList.add('csl-entry');
|
||||
var biblioDiv = window.document.getElementById('ref-' + cite);
|
||||
if (biblioDiv) {
|
||||
citeDiv.innerHTML = biblioDiv.innerHTML;
|
||||
}
|
||||
popup.appendChild(citeDiv);
|
||||
});
|
||||
return popup.innerHTML;
|
||||
});
|
||||
}
|
||||
}
|
||||
});
|
||||
</script>
|
||||
</div> <!-- /content -->
|
||||
|
||||
|
||||
|
||||
|
||||
</body></html>
|
||||
837
docs/api/loaders.processor.html
Normal file
837
docs/api/loaders.processor.html
Normal file
@@ -0,0 +1,837 @@
|
||||
<!DOCTYPE html>
|
||||
<html xmlns="http://www.w3.org/1999/xhtml" lang="en" xml:lang="en"><head>
|
||||
|
||||
<meta charset="utf-8">
|
||||
<meta name="generator" content="quarto-1.7.31">
|
||||
|
||||
<meta name="viewport" content="width=device-width, initial-scale=1.0, user-scalable=yes">
|
||||
|
||||
|
||||
<title>loaders.processor – Axolotl</title>
|
||||
<style>
|
||||
code{white-space: pre-wrap;}
|
||||
span.smallcaps{font-variant: small-caps;}
|
||||
div.columns{display: flex; gap: min(4vw, 1.5em);}
|
||||
div.column{flex: auto; overflow-x: auto;}
|
||||
div.hanging-indent{margin-left: 1.5em; text-indent: -1.5em;}
|
||||
ul.task-list{list-style: none;}
|
||||
ul.task-list li input[type="checkbox"] {
|
||||
width: 0.8em;
|
||||
margin: 0 0.8em 0.2em -1em; /* quarto-specific, see https://github.com/quarto-dev/quarto-cli/issues/4556 */
|
||||
vertical-align: middle;
|
||||
}
|
||||
</style>
|
||||
|
||||
|
||||
<script src="../../site_libs/quarto-nav/quarto-nav.js"></script>
|
||||
<script src="../../site_libs/clipboard/clipboard.min.js"></script>
|
||||
<script src="../../site_libs/quarto-search/autocomplete.umd.js"></script>
|
||||
<script src="../../site_libs/quarto-search/fuse.min.js"></script>
|
||||
<script src="../../site_libs/quarto-search/quarto-search.js"></script>
|
||||
<meta name="quarto:offset" content="../../">
|
||||
<link href="../../favicon.jpg" rel="icon" type="image/jpeg">
|
||||
<script src="../../site_libs/quarto-html/quarto.js" type="module"></script>
|
||||
<script src="../../site_libs/quarto-html/tabsets/tabsets.js" type="module"></script>
|
||||
<script src="../../site_libs/quarto-html/popper.min.js"></script>
|
||||
<script src="../../site_libs/quarto-html/tippy.umd.min.js"></script>
|
||||
<script src="../../site_libs/quarto-html/anchor.min.js"></script>
|
||||
<link href="../../site_libs/quarto-html/tippy.css" rel="stylesheet">
|
||||
<link href="../../site_libs/quarto-html/quarto-syntax-highlighting-dark-8ef56b68f8fa1e9d2ba328e99e439f80.css" rel="stylesheet" id="quarto-text-highlighting-styles">
|
||||
<script src="../../site_libs/bootstrap/bootstrap.min.js"></script>
|
||||
<link href="../../site_libs/bootstrap/bootstrap-icons.css" rel="stylesheet">
|
||||
<link href="../../site_libs/bootstrap/bootstrap-2288ecdcbf81d2ab6432743cedd71d9a.min.css" rel="stylesheet" append-hash="true" id="quarto-bootstrap" data-mode="dark">
|
||||
<script id="quarto-search-options" type="application/json">{
|
||||
"location": "navbar",
|
||||
"copy-button": false,
|
||||
"collapse-after": 3,
|
||||
"panel-placement": "end",
|
||||
"type": "overlay",
|
||||
"limit": 50,
|
||||
"keyboard-shortcut": [
|
||||
"f",
|
||||
"/",
|
||||
"s"
|
||||
],
|
||||
"show-item-context": false,
|
||||
"language": {
|
||||
"search-no-results-text": "No results",
|
||||
"search-matching-documents-text": "matching documents",
|
||||
"search-copy-link-title": "Copy link to search",
|
||||
"search-hide-matches-text": "Hide additional matches",
|
||||
"search-more-match-text": "more match in this document",
|
||||
"search-more-matches-text": "more matches in this document",
|
||||
"search-clear-button-title": "Clear",
|
||||
"search-text-placeholder": "",
|
||||
"search-detached-cancel-button-title": "Cancel",
|
||||
"search-submit-button-title": "Submit",
|
||||
"search-label": "Search"
|
||||
}
|
||||
}</script>
|
||||
|
||||
|
||||
<link rel="stylesheet" href="../../styles.css">
|
||||
</head>
|
||||
|
||||
<body class="nav-sidebar docked nav-fixed quarto-light">
|
||||
|
||||
<div id="quarto-search-results"></div>
|
||||
<header id="quarto-header" class="headroom fixed-top">
|
||||
<nav class="navbar navbar-expand " data-bs-theme="dark">
|
||||
<div class="navbar-container container-fluid">
|
||||
<div class="navbar-brand-container mx-auto">
|
||||
<a href="../../index.html" class="navbar-brand navbar-brand-logo">
|
||||
<img src="../../image/axolotl_logo_digital_white.svg" alt="" class="navbar-logo">
|
||||
</a>
|
||||
</div>
|
||||
<div class="quarto-navbar-tools tools-wide tools-end">
|
||||
<a href="https://twitter.com/axolotl_ai" title="" class="quarto-navigation-tool px-1" aria-label=""><i class="bi bi-twitter"></i></a>
|
||||
<a href="https://github.com/axolotl-ai-cloud/axolotl/" title="" class="quarto-navigation-tool px-1" aria-label=""><i class="bi bi-github"></i></a>
|
||||
<a href="https://discord.gg/7m9sfhzaf3" title="" class="quarto-navigation-tool px-1" aria-label=""><i class="bi bi-discord"></i></a>
|
||||
</div>
|
||||
<div id="quarto-search" class="" title="Search"></div>
|
||||
</div> <!-- /container-fluid -->
|
||||
</nav>
|
||||
<nav class="quarto-secondary-nav">
|
||||
<div class="container-fluid d-flex">
|
||||
<button type="button" class="quarto-btn-toggle btn" data-bs-toggle="collapse" role="button" data-bs-target=".quarto-sidebar-collapse-item" aria-controls="quarto-sidebar" aria-expanded="false" aria-label="Toggle sidebar navigation" onclick="if (window.quartoToggleHeadroom) { window.quartoToggleHeadroom(); }">
|
||||
<i class="bi bi-layout-text-sidebar-reverse"></i>
|
||||
</button>
|
||||
<nav class="quarto-page-breadcrumbs" aria-label="breadcrumb"><ol class="breadcrumb"></ol></nav>
|
||||
<a class="flex-grow-1" role="navigation" data-bs-toggle="collapse" data-bs-target=".quarto-sidebar-collapse-item" aria-controls="quarto-sidebar" aria-expanded="false" aria-label="Toggle sidebar navigation" onclick="if (window.quartoToggleHeadroom) { window.quartoToggleHeadroom(); }">
|
||||
</a>
|
||||
</div>
|
||||
</nav>
|
||||
</header>
|
||||
<!-- content -->
|
||||
<div id="quarto-content" class="quarto-container page-columns page-rows-contents page-layout-article page-navbar">
|
||||
<!-- sidebar -->
|
||||
<nav id="quarto-sidebar" class="sidebar collapse collapse-horizontal quarto-sidebar-collapse-item sidebar-navigation docked overflow-auto">
|
||||
<div class="sidebar-menu-container">
|
||||
<ul class="list-unstyled mt-1">
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../index.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Home</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item sidebar-item-section">
|
||||
<div class="sidebar-item-container">
|
||||
<a class="sidebar-item-text sidebar-link text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-1" role="navigation" aria-expanded="true">
|
||||
<span class="menu-text">Getting Started</span></a>
|
||||
<a class="sidebar-item-toggle text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-1" role="navigation" aria-expanded="true" aria-label="Toggle section">
|
||||
<i class="bi bi-chevron-right ms-2"></i>
|
||||
</a>
|
||||
</div>
|
||||
<ul id="quarto-sidebar-section-1" class="collapse list-unstyled sidebar-section depth1 show">
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/getting-started.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Quickstart</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/installation.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Installation</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/inference.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Inference and Merging</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/cli.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Command Line Interface (CLI)</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/config.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Config Reference</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/api" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">API Reference</span></a>
|
||||
</div>
|
||||
</li>
|
||||
</ul>
|
||||
</li>
|
||||
<li class="sidebar-item sidebar-item-section">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/dataset-formats/index.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Dataset Formats</span></a>
|
||||
<a class="sidebar-item-toggle text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-2" role="navigation" aria-expanded="true" aria-label="Toggle section">
|
||||
<i class="bi bi-chevron-right ms-2"></i>
|
||||
</a>
|
||||
</div>
|
||||
<ul id="quarto-sidebar-section-2" class="collapse list-unstyled sidebar-section depth1 show">
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/dataset-formats/pretraining.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Pre-training</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/dataset-formats/inst_tune.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Instruction Tuning</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/dataset-formats/conversation.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Conversation</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/dataset-formats/stepwise_supervised.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Stepwise Supervised Format</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/dataset-formats/template_free.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Template-Free</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/dataset-formats/tokenized.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Custom Pre-Tokenized Dataset</span></a>
|
||||
</div>
|
||||
</li>
|
||||
</ul>
|
||||
</li>
|
||||
<li class="sidebar-item sidebar-item-section">
|
||||
<div class="sidebar-item-container">
|
||||
<a class="sidebar-item-text sidebar-link text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-3" role="navigation" aria-expanded="true">
|
||||
<span class="menu-text">Deployments</span></a>
|
||||
<a class="sidebar-item-toggle text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-3" role="navigation" aria-expanded="true" aria-label="Toggle section">
|
||||
<i class="bi bi-chevron-right ms-2"></i>
|
||||
</a>
|
||||
</div>
|
||||
<ul id="quarto-sidebar-section-3" class="collapse list-unstyled sidebar-section depth1 show">
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/docker.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Docker</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/multi-gpu.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Multi-GPU</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/multi-node.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Multi Node</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/ray-integration.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Ray Train</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/amd_hpc.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">AMD GPUs on HPC Systems</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/mac.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Mac M-series</span></a>
|
||||
</div>
|
||||
</li>
|
||||
</ul>
|
||||
</li>
|
||||
<li class="sidebar-item sidebar-item-section">
|
||||
<div class="sidebar-item-container">
|
||||
<a class="sidebar-item-text sidebar-link text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-4" role="navigation" aria-expanded="true">
|
||||
<span class="menu-text">How To Guides</span></a>
|
||||
<a class="sidebar-item-toggle text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-4" role="navigation" aria-expanded="true" aria-label="Toggle section">
|
||||
<i class="bi bi-chevron-right ms-2"></i>
|
||||
</a>
|
||||
</div>
|
||||
<ul id="quarto-sidebar-section-4" class="collapse list-unstyled sidebar-section depth1 show">
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/multimodal.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">MultiModal / Vision Language Models (BETA)</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/rlhf.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">RLHF (Beta)</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/reward_modelling.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Reward Modelling</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/lr_groups.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Learning Rate Groups</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/lora_optims.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">LoRA Optimizations</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/dataset_loading.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Dataset Loading</span></a>
|
||||
</div>
|
||||
</li>
|
||||
</ul>
|
||||
</li>
|
||||
<li class="sidebar-item sidebar-item-section">
|
||||
<div class="sidebar-item-container">
|
||||
<a class="sidebar-item-text sidebar-link text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-5" role="navigation" aria-expanded="true">
|
||||
<span class="menu-text">Core Concepts</span></a>
|
||||
<a class="sidebar-item-toggle text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-5" role="navigation" aria-expanded="true" aria-label="Toggle section">
|
||||
<i class="bi bi-chevron-right ms-2"></i>
|
||||
</a>
|
||||
</div>
|
||||
<ul id="quarto-sidebar-section-5" class="collapse list-unstyled sidebar-section depth1 show">
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/batch_vs_grad.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Batch size vs Gradient accumulation</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/dataset_preprocessing.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Dataset Preprocessing</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/multipack.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Multipack (Sample Packing)</span></a>
|
||||
</div>
|
||||
</li>
|
||||
</ul>
|
||||
</li>
|
||||
<li class="sidebar-item sidebar-item-section">
|
||||
<div class="sidebar-item-container">
|
||||
<a class="sidebar-item-text sidebar-link text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-6" role="navigation" aria-expanded="true">
|
||||
<span class="menu-text">Advanced Features</span></a>
|
||||
<a class="sidebar-item-toggle text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-6" role="navigation" aria-expanded="true" aria-label="Toggle section">
|
||||
<i class="bi bi-chevron-right ms-2"></i>
|
||||
</a>
|
||||
</div>
|
||||
<ul id="quarto-sidebar-section-6" class="collapse list-unstyled sidebar-section depth1 show">
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/fsdp_qlora.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">FDSP + QLoRA</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/unsloth.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Unsloth</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/torchao.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">PyTorch ao</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/custom_integrations.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Custom Integrations</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/sequence_parallelism.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Sequence Parallelism</span></a>
|
||||
</div>
|
||||
</li>
|
||||
</ul>
|
||||
</li>
|
||||
<li class="sidebar-item sidebar-item-section">
|
||||
<div class="sidebar-item-container">
|
||||
<a class="sidebar-item-text sidebar-link text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-7" role="navigation" aria-expanded="true">
|
||||
<span class="menu-text">Troubleshooting</span></a>
|
||||
<a class="sidebar-item-toggle text-start" data-bs-toggle="collapse" data-bs-target="#quarto-sidebar-section-7" role="navigation" aria-expanded="true" aria-label="Toggle section">
|
||||
<i class="bi bi-chevron-right ms-2"></i>
|
||||
</a>
|
||||
</div>
|
||||
<ul id="quarto-sidebar-section-7" class="collapse list-unstyled sidebar-section depth1 show">
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/faq.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">FAQ</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/debugging.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">Debugging</span></a>
|
||||
</div>
|
||||
</li>
|
||||
<li class="sidebar-item">
|
||||
<div class="sidebar-item-container">
|
||||
<a href="../../docs/nccl.html" class="sidebar-item-text sidebar-link">
|
||||
<span class="menu-text">NCCL</span></a>
|
||||
</div>
|
||||
</li>
|
||||
</ul>
|
||||
</li>
|
||||
</ul>
|
||||
</div>
|
||||
</nav>
|
||||
<div id="quarto-sidebar-glass" class="quarto-sidebar-collapse-item" data-bs-toggle="collapse" data-bs-target=".quarto-sidebar-collapse-item"></div>
|
||||
<!-- margin-sidebar -->
|
||||
<div id="quarto-margin-sidebar" class="sidebar margin-sidebar">
|
||||
<nav id="TOC" role="doc-toc" class="toc-active">
|
||||
<h2 id="toc-title">On this page</h2>
|
||||
|
||||
<ul>
|
||||
<li><a href="#axolotl.loaders.processor" id="toc-axolotl.loaders.processor" class="nav-link active" data-scroll-target="#axolotl.loaders.processor">loaders.processor</a></li>
|
||||
</ul>
|
||||
</nav>
|
||||
</div>
|
||||
<!-- main -->
|
||||
<main class="content" id="quarto-document-content"><header id="title-block-header" class="quarto-title-block"></header>
|
||||
|
||||
|
||||
|
||||
|
||||
<section id="axolotl.loaders.processor" class="level1">
|
||||
<h1>loaders.processor</h1>
|
||||
<p><code>loaders.processor</code></p>
|
||||
<p>Processor loading functionality for multi-modal models</p>
|
||||
|
||||
|
||||
</section>
|
||||
|
||||
</main> <!-- /main -->
|
||||
<script id="quarto-html-after-body" type="application/javascript">
|
||||
window.document.addEventListener("DOMContentLoaded", function (event) {
|
||||
const icon = "";
|
||||
const anchorJS = new window.AnchorJS();
|
||||
anchorJS.options = {
|
||||
placement: 'right',
|
||||
icon: icon
|
||||
};
|
||||
anchorJS.add('.anchored');
|
||||
const isCodeAnnotation = (el) => {
|
||||
for (const clz of el.classList) {
|
||||
if (clz.startsWith('code-annotation-')) {
|
||||
return true;
|
||||
}
|
||||
}
|
||||
return false;
|
||||
}
|
||||
const onCopySuccess = function(e) {
|
||||
// button target
|
||||
const button = e.trigger;
|
||||
// don't keep focus
|
||||
button.blur();
|
||||
// flash "checked"
|
||||
button.classList.add('code-copy-button-checked');
|
||||
var currentTitle = button.getAttribute("title");
|
||||
button.setAttribute("title", "Copied!");
|
||||
let tooltip;
|
||||
if (window.bootstrap) {
|
||||
button.setAttribute("data-bs-toggle", "tooltip");
|
||||
button.setAttribute("data-bs-placement", "left");
|
||||
button.setAttribute("data-bs-title", "Copied!");
|
||||
tooltip = new bootstrap.Tooltip(button,
|
||||
{ trigger: "manual",
|
||||
customClass: "code-copy-button-tooltip",
|
||||
offset: [0, -8]});
|
||||
tooltip.show();
|
||||
}
|
||||
setTimeout(function() {
|
||||
if (tooltip) {
|
||||
tooltip.hide();
|
||||
button.removeAttribute("data-bs-title");
|
||||
button.removeAttribute("data-bs-toggle");
|
||||
button.removeAttribute("data-bs-placement");
|
||||
}
|
||||
button.setAttribute("title", currentTitle);
|
||||
button.classList.remove('code-copy-button-checked');
|
||||
}, 1000);
|
||||
// clear code selection
|
||||
e.clearSelection();
|
||||
}
|
||||
const getTextToCopy = function(trigger) {
|
||||
const codeEl = trigger.previousElementSibling.cloneNode(true);
|
||||
for (const childEl of codeEl.children) {
|
||||
if (isCodeAnnotation(childEl)) {
|
||||
childEl.remove();
|
||||
}
|
||||
}
|
||||
return codeEl.innerText;
|
||||
}
|
||||
const clipboard = new window.ClipboardJS('.code-copy-button:not([data-in-quarto-modal])', {
|
||||
text: getTextToCopy
|
||||
});
|
||||
clipboard.on('success', onCopySuccess);
|
||||
if (window.document.getElementById('quarto-embedded-source-code-modal')) {
|
||||
const clipboardModal = new window.ClipboardJS('.code-copy-button[data-in-quarto-modal]', {
|
||||
text: getTextToCopy,
|
||||
container: window.document.getElementById('quarto-embedded-source-code-modal')
|
||||
});
|
||||
clipboardModal.on('success', onCopySuccess);
|
||||
}
|
||||
var localhostRegex = new RegExp(/^(?:http|https):\/\/localhost\:?[0-9]*\//);
|
||||
var mailtoRegex = new RegExp(/^mailto:/);
|
||||
var filterRegex = new RegExp("https:\/\/docs\.axolotl\.ai");
|
||||
var isInternal = (href) => {
|
||||
return filterRegex.test(href) || localhostRegex.test(href) || mailtoRegex.test(href);
|
||||
}
|
||||
// Inspect non-navigation links and adorn them if external
|
||||
var links = window.document.querySelectorAll('a[href]:not(.nav-link):not(.navbar-brand):not(.toc-action):not(.sidebar-link):not(.sidebar-item-toggle):not(.pagination-link):not(.no-external):not([aria-hidden]):not(.dropdown-item):not(.quarto-navigation-tool):not(.about-link)');
|
||||
for (var i=0; i<links.length; i++) {
|
||||
const link = links[i];
|
||||
if (!isInternal(link.href)) {
|
||||
// undo the damage that might have been done by quarto-nav.js in the case of
|
||||
// links that we want to consider external
|
||||
if (link.dataset.originalHref !== undefined) {
|
||||
link.href = link.dataset.originalHref;
|
||||
}
|
||||
}
|
||||
}
|
||||
function tippyHover(el, contentFn, onTriggerFn, onUntriggerFn) {
|
||||
const config = {
|
||||
allowHTML: true,
|
||||
maxWidth: 500,
|
||||
delay: 100,
|
||||
arrow: false,
|
||||
appendTo: function(el) {
|
||||
return el.parentElement;
|
||||
},
|
||||
interactive: true,
|
||||
interactiveBorder: 10,
|
||||
theme: 'quarto',
|
||||
placement: 'bottom-start',
|
||||
};
|
||||
if (contentFn) {
|
||||
config.content = contentFn;
|
||||
}
|
||||
if (onTriggerFn) {
|
||||
config.onTrigger = onTriggerFn;
|
||||
}
|
||||
if (onUntriggerFn) {
|
||||
config.onUntrigger = onUntriggerFn;
|
||||
}
|
||||
window.tippy(el, config);
|
||||
}
|
||||
const noterefs = window.document.querySelectorAll('a[role="doc-noteref"]');
|
||||
for (var i=0; i<noterefs.length; i++) {
|
||||
const ref = noterefs[i];
|
||||
tippyHover(ref, function() {
|
||||
// use id or data attribute instead here
|
||||
let href = ref.getAttribute('data-footnote-href') || ref.getAttribute('href');
|
||||
try { href = new URL(href).hash; } catch {}
|
||||
const id = href.replace(/^#\/?/, "");
|
||||
const note = window.document.getElementById(id);
|
||||
if (note) {
|
||||
return note.innerHTML;
|
||||
} else {
|
||||
return "";
|
||||
}
|
||||
});
|
||||
}
|
||||
const xrefs = window.document.querySelectorAll('a.quarto-xref');
|
||||
const processXRef = (id, note) => {
|
||||
// Strip column container classes
|
||||
const stripColumnClz = (el) => {
|
||||
el.classList.remove("page-full", "page-columns");
|
||||
if (el.children) {
|
||||
for (const child of el.children) {
|
||||
stripColumnClz(child);
|
||||
}
|
||||
}
|
||||
}
|
||||
stripColumnClz(note)
|
||||
if (id === null || id.startsWith('sec-')) {
|
||||
// Special case sections, only their first couple elements
|
||||
const container = document.createElement("div");
|
||||
if (note.children && note.children.length > 2) {
|
||||
container.appendChild(note.children[0].cloneNode(true));
|
||||
for (let i = 1; i < note.children.length; i++) {
|
||||
const child = note.children[i];
|
||||
if (child.tagName === "P" && child.innerText === "") {
|
||||
continue;
|
||||
} else {
|
||||
container.appendChild(child.cloneNode(true));
|
||||
break;
|
||||
}
|
||||
}
|
||||
if (window.Quarto?.typesetMath) {
|
||||
window.Quarto.typesetMath(container);
|
||||
}
|
||||
return container.innerHTML
|
||||
} else {
|
||||
if (window.Quarto?.typesetMath) {
|
||||
window.Quarto.typesetMath(note);
|
||||
}
|
||||
return note.innerHTML;
|
||||
}
|
||||
} else {
|
||||
// Remove any anchor links if they are present
|
||||
const anchorLink = note.querySelector('a.anchorjs-link');
|
||||
if (anchorLink) {
|
||||
anchorLink.remove();
|
||||
}
|
||||
if (window.Quarto?.typesetMath) {
|
||||
window.Quarto.typesetMath(note);
|
||||
}
|
||||
if (note.classList.contains("callout")) {
|
||||
return note.outerHTML;
|
||||
} else {
|
||||
return note.innerHTML;
|
||||
}
|
||||
}
|
||||
}
|
||||
for (var i=0; i<xrefs.length; i++) {
|
||||
const xref = xrefs[i];
|
||||
tippyHover(xref, undefined, function(instance) {
|
||||
instance.disable();
|
||||
let url = xref.getAttribute('href');
|
||||
let hash = undefined;
|
||||
if (url.startsWith('#')) {
|
||||
hash = url;
|
||||
} else {
|
||||
try { hash = new URL(url).hash; } catch {}
|
||||
}
|
||||
if (hash) {
|
||||
const id = hash.replace(/^#\/?/, "");
|
||||
const note = window.document.getElementById(id);
|
||||
if (note !== null) {
|
||||
try {
|
||||
const html = processXRef(id, note.cloneNode(true));
|
||||
instance.setContent(html);
|
||||
} finally {
|
||||
instance.enable();
|
||||
instance.show();
|
||||
}
|
||||
} else {
|
||||
// See if we can fetch this
|
||||
fetch(url.split('#')[0])
|
||||
.then(res => res.text())
|
||||
.then(html => {
|
||||
const parser = new DOMParser();
|
||||
const htmlDoc = parser.parseFromString(html, "text/html");
|
||||
const note = htmlDoc.getElementById(id);
|
||||
if (note !== null) {
|
||||
const html = processXRef(id, note);
|
||||
instance.setContent(html);
|
||||
}
|
||||
}).finally(() => {
|
||||
instance.enable();
|
||||
instance.show();
|
||||
});
|
||||
}
|
||||
} else {
|
||||
// See if we can fetch a full url (with no hash to target)
|
||||
// This is a special case and we should probably do some content thinning / targeting
|
||||
fetch(url)
|
||||
.then(res => res.text())
|
||||
.then(html => {
|
||||
const parser = new DOMParser();
|
||||
const htmlDoc = parser.parseFromString(html, "text/html");
|
||||
const note = htmlDoc.querySelector('main.content');
|
||||
if (note !== null) {
|
||||
// This should only happen for chapter cross references
|
||||
// (since there is no id in the URL)
|
||||
// remove the first header
|
||||
if (note.children.length > 0 && note.children[0].tagName === "HEADER") {
|
||||
note.children[0].remove();
|
||||
}
|
||||
const html = processXRef(null, note);
|
||||
instance.setContent(html);
|
||||
}
|
||||
}).finally(() => {
|
||||
instance.enable();
|
||||
instance.show();
|
||||
});
|
||||
}
|
||||
}, function(instance) {
|
||||
});
|
||||
}
|
||||
let selectedAnnoteEl;
|
||||
const selectorForAnnotation = ( cell, annotation) => {
|
||||
let cellAttr = 'data-code-cell="' + cell + '"';
|
||||
let lineAttr = 'data-code-annotation="' + annotation + '"';
|
||||
const selector = 'span[' + cellAttr + '][' + lineAttr + ']';
|
||||
return selector;
|
||||
}
|
||||
const selectCodeLines = (annoteEl) => {
|
||||
const doc = window.document;
|
||||
const targetCell = annoteEl.getAttribute("data-target-cell");
|
||||
const targetAnnotation = annoteEl.getAttribute("data-target-annotation");
|
||||
const annoteSpan = window.document.querySelector(selectorForAnnotation(targetCell, targetAnnotation));
|
||||
const lines = annoteSpan.getAttribute("data-code-lines").split(",");
|
||||
const lineIds = lines.map((line) => {
|
||||
return targetCell + "-" + line;
|
||||
})
|
||||
let top = null;
|
||||
let height = null;
|
||||
let parent = null;
|
||||
if (lineIds.length > 0) {
|
||||
//compute the position of the single el (top and bottom and make a div)
|
||||
const el = window.document.getElementById(lineIds[0]);
|
||||
top = el.offsetTop;
|
||||
height = el.offsetHeight;
|
||||
parent = el.parentElement.parentElement;
|
||||
if (lineIds.length > 1) {
|
||||
const lastEl = window.document.getElementById(lineIds[lineIds.length - 1]);
|
||||
const bottom = lastEl.offsetTop + lastEl.offsetHeight;
|
||||
height = bottom - top;
|
||||
}
|
||||
if (top !== null && height !== null && parent !== null) {
|
||||
// cook up a div (if necessary) and position it
|
||||
let div = window.document.getElementById("code-annotation-line-highlight");
|
||||
if (div === null) {
|
||||
div = window.document.createElement("div");
|
||||
div.setAttribute("id", "code-annotation-line-highlight");
|
||||
div.style.position = 'absolute';
|
||||
parent.appendChild(div);
|
||||
}
|
||||
div.style.top = top - 2 + "px";
|
||||
div.style.height = height + 4 + "px";
|
||||
div.style.left = 0;
|
||||
let gutterDiv = window.document.getElementById("code-annotation-line-highlight-gutter");
|
||||
if (gutterDiv === null) {
|
||||
gutterDiv = window.document.createElement("div");
|
||||
gutterDiv.setAttribute("id", "code-annotation-line-highlight-gutter");
|
||||
gutterDiv.style.position = 'absolute';
|
||||
const codeCell = window.document.getElementById(targetCell);
|
||||
const gutter = codeCell.querySelector('.code-annotation-gutter');
|
||||
gutter.appendChild(gutterDiv);
|
||||
}
|
||||
gutterDiv.style.top = top - 2 + "px";
|
||||
gutterDiv.style.height = height + 4 + "px";
|
||||
}
|
||||
selectedAnnoteEl = annoteEl;
|
||||
}
|
||||
};
|
||||
const unselectCodeLines = () => {
|
||||
const elementsIds = ["code-annotation-line-highlight", "code-annotation-line-highlight-gutter"];
|
||||
elementsIds.forEach((elId) => {
|
||||
const div = window.document.getElementById(elId);
|
||||
if (div) {
|
||||
div.remove();
|
||||
}
|
||||
});
|
||||
selectedAnnoteEl = undefined;
|
||||
};
|
||||
// Handle positioning of the toggle
|
||||
window.addEventListener(
|
||||
"resize",
|
||||
throttle(() => {
|
||||
elRect = undefined;
|
||||
if (selectedAnnoteEl) {
|
||||
selectCodeLines(selectedAnnoteEl);
|
||||
}
|
||||
}, 10)
|
||||
);
|
||||
function throttle(fn, ms) {
|
||||
let throttle = false;
|
||||
let timer;
|
||||
return (...args) => {
|
||||
if(!throttle) { // first call gets through
|
||||
fn.apply(this, args);
|
||||
throttle = true;
|
||||
} else { // all the others get throttled
|
||||
if(timer) clearTimeout(timer); // cancel #2
|
||||
timer = setTimeout(() => {
|
||||
fn.apply(this, args);
|
||||
timer = throttle = false;
|
||||
}, ms);
|
||||
}
|
||||
};
|
||||
}
|
||||
// Attach click handler to the DT
|
||||
const annoteDls = window.document.querySelectorAll('dt[data-target-cell]');
|
||||
for (const annoteDlNode of annoteDls) {
|
||||
annoteDlNode.addEventListener('click', (event) => {
|
||||
const clickedEl = event.target;
|
||||
if (clickedEl !== selectedAnnoteEl) {
|
||||
unselectCodeLines();
|
||||
const activeEl = window.document.querySelector('dt[data-target-cell].code-annotation-active');
|
||||
if (activeEl) {
|
||||
activeEl.classList.remove('code-annotation-active');
|
||||
}
|
||||
selectCodeLines(clickedEl);
|
||||
clickedEl.classList.add('code-annotation-active');
|
||||
} else {
|
||||
// Unselect the line
|
||||
unselectCodeLines();
|
||||
clickedEl.classList.remove('code-annotation-active');
|
||||
}
|
||||
});
|
||||
}
|
||||
const findCites = (el) => {
|
||||
const parentEl = el.parentElement;
|
||||
if (parentEl) {
|
||||
const cites = parentEl.dataset.cites;
|
||||
if (cites) {
|
||||
return {
|
||||
el,
|
||||
cites: cites.split(' ')
|
||||
};
|
||||
} else {
|
||||
return findCites(el.parentElement)
|
||||
}
|
||||
} else {
|
||||
return undefined;
|
||||
}
|
||||
};
|
||||
var bibliorefs = window.document.querySelectorAll('a[role="doc-biblioref"]');
|
||||
for (var i=0; i<bibliorefs.length; i++) {
|
||||
const ref = bibliorefs[i];
|
||||
const citeInfo = findCites(ref);
|
||||
if (citeInfo) {
|
||||
tippyHover(citeInfo.el, function() {
|
||||
var popup = window.document.createElement('div');
|
||||
citeInfo.cites.forEach(function(cite) {
|
||||
var citeDiv = window.document.createElement('div');
|
||||
citeDiv.classList.add('hanging-indent');
|
||||
citeDiv.classList.add('csl-entry');
|
||||
var biblioDiv = window.document.getElementById('ref-' + cite);
|
||||
if (biblioDiv) {
|
||||
citeDiv.innerHTML = biblioDiv.innerHTML;
|
||||
}
|
||||
popup.appendChild(citeDiv);
|
||||
});
|
||||
return popup.innerHTML;
|
||||
});
|
||||
}
|
||||
}
|
||||
});
|
||||
</script>
|
||||
</div> <!-- /content -->
|
||||
|
||||
|
||||
|
||||
|
||||
</body></html>
|
||||
@@ -7,7 +7,7 @@
|
||||
<meta name="viewport" content="width=device-width, initial-scale=1.0, user-scalable=yes">
|
||||
|
||||
|
||||
<title>utils.models – Axolotl</title>
|
||||
<title>loaders.tokenizer – Axolotl</title>
|
||||
<style>
|
||||
code{white-space: pre-wrap;}
|
||||
span.smallcaps{font-variant: small-caps;}
|
||||
@@ -446,20 +446,12 @@ pre > code.sourceCode > span > a:first-child::before { text-decoration: underlin
|
||||
<h2 id="toc-title">On this page</h2>
|
||||
|
||||
<ul>
|
||||
<li><a href="#axolotl.utils.models" id="toc-axolotl.utils.models" class="nav-link active" data-scroll-target="#axolotl.utils.models">utils.models</a>
|
||||
<li><a href="#axolotl.loaders.tokenizer" id="toc-axolotl.loaders.tokenizer" class="nav-link active" data-scroll-target="#axolotl.loaders.tokenizer">loaders.tokenizer</a>
|
||||
<ul class="collapse">
|
||||
<li><a href="#classes" id="toc-classes" class="nav-link" data-scroll-target="#classes">Classes</a>
|
||||
<ul class="collapse">
|
||||
<li><a href="#axolotl.utils.models.ModelLoader" id="toc-axolotl.utils.models.ModelLoader" class="nav-link" data-scroll-target="#axolotl.utils.models.ModelLoader">ModelLoader</a></li>
|
||||
</ul></li>
|
||||
<li><a href="#functions" id="toc-functions" class="nav-link" data-scroll-target="#functions">Functions</a>
|
||||
<ul class="collapse">
|
||||
<li><a href="#axolotl.utils.models.get_module_class_from_name" id="toc-axolotl.utils.models.get_module_class_from_name" class="nav-link" data-scroll-target="#axolotl.utils.models.get_module_class_from_name">get_module_class_from_name</a></li>
|
||||
<li><a href="#axolotl.utils.models.load_model" id="toc-axolotl.utils.models.load_model" class="nav-link" data-scroll-target="#axolotl.utils.models.load_model">load_model</a></li>
|
||||
<li><a href="#axolotl.utils.models.load_tokenizer" id="toc-axolotl.utils.models.load_tokenizer" class="nav-link" data-scroll-target="#axolotl.utils.models.load_tokenizer">load_tokenizer</a></li>
|
||||
<li><a href="#axolotl.utils.models.modify_tokenizer_files" id="toc-axolotl.utils.models.modify_tokenizer_files" class="nav-link" data-scroll-target="#axolotl.utils.models.modify_tokenizer_files">modify_tokenizer_files</a></li>
|
||||
<li><a href="#axolotl.utils.models.setup_quantized_meta_for_peft" id="toc-axolotl.utils.models.setup_quantized_meta_for_peft" class="nav-link" data-scroll-target="#axolotl.utils.models.setup_quantized_meta_for_peft">setup_quantized_meta_for_peft</a></li>
|
||||
<li><a href="#axolotl.utils.models.setup_quantized_peft_meta_for_training" id="toc-axolotl.utils.models.setup_quantized_peft_meta_for_training" class="nav-link" data-scroll-target="#axolotl.utils.models.setup_quantized_peft_meta_for_training">setup_quantized_peft_meta_for_training</a></li>
|
||||
<li><a href="#axolotl.loaders.tokenizer.load_tokenizer" id="toc-axolotl.loaders.tokenizer.load_tokenizer" class="nav-link" data-scroll-target="#axolotl.loaders.tokenizer.load_tokenizer">load_tokenizer</a></li>
|
||||
<li><a href="#axolotl.loaders.tokenizer.modify_tokenizer_files" id="toc-axolotl.loaders.tokenizer.modify_tokenizer_files" class="nav-link" data-scroll-target="#axolotl.loaders.tokenizer.modify_tokenizer_files">modify_tokenizer_files</a></li>
|
||||
</ul></li>
|
||||
</ul></li>
|
||||
</ul>
|
||||
@@ -471,109 +463,10 @@ pre > code.sourceCode > span > a:first-child::before { text-decoration: underlin
|
||||
|
||||
|
||||
|
||||
<section id="axolotl.utils.models" class="level1">
|
||||
<h1>utils.models</h1>
|
||||
<p><code>utils.models</code></p>
|
||||
<p>Module for models and model loading</p>
|
||||
<section id="classes" class="level2">
|
||||
<h2 class="anchored" data-anchor-id="classes">Classes</h2>
|
||||
<table class="caption-top table">
|
||||
<thead>
|
||||
<tr class="header">
|
||||
<th>Name</th>
|
||||
<th>Description</th>
|
||||
</tr>
|
||||
</thead>
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td><a href="#axolotl.utils.models.ModelLoader">ModelLoader</a></td>
|
||||
<td>ModelLoader: managing all the config and monkey patches while loading model</td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
<section id="axolotl.utils.models.ModelLoader" class="level3">
|
||||
<h3 class="anchored" data-anchor-id="axolotl.utils.models.ModelLoader">ModelLoader</h3>
|
||||
<div class="sourceCode" id="cb1"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb1-1"><a href="#cb1-1" aria-hidden="true" tabindex="-1"></a>utils.models.ModelLoader(</span>
|
||||
<span id="cb1-2"><a href="#cb1-2" aria-hidden="true" tabindex="-1"></a> <span class="va">self</span>,</span>
|
||||
<span id="cb1-3"><a href="#cb1-3" aria-hidden="true" tabindex="-1"></a> cfg,</span>
|
||||
<span id="cb1-4"><a href="#cb1-4" aria-hidden="true" tabindex="-1"></a> tokenizer,</span>
|
||||
<span id="cb1-5"><a href="#cb1-5" aria-hidden="true" tabindex="-1"></a> <span class="op">*</span>,</span>
|
||||
<span id="cb1-6"><a href="#cb1-6" aria-hidden="true" tabindex="-1"></a> processor<span class="op">=</span><span class="va">None</span>,</span>
|
||||
<span id="cb1-7"><a href="#cb1-7" aria-hidden="true" tabindex="-1"></a> inference<span class="op">=</span><span class="va">False</span>,</span>
|
||||
<span id="cb1-8"><a href="#cb1-8" aria-hidden="true" tabindex="-1"></a> reference_model<span class="op">=</span><span class="va">False</span>,</span>
|
||||
<span id="cb1-9"><a href="#cb1-9" aria-hidden="true" tabindex="-1"></a> <span class="op">**</span>kwargs,</span>
|
||||
<span id="cb1-10"><a href="#cb1-10" aria-hidden="true" tabindex="-1"></a>)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>ModelLoader: managing all the config and monkey patches while loading model</p>
|
||||
<section id="attributes" class="level4">
|
||||
<h4 class="anchored" data-anchor-id="attributes">Attributes</h4>
|
||||
<table class="caption-top table">
|
||||
<thead>
|
||||
<tr class="header">
|
||||
<th>Name</th>
|
||||
<th>Description</th>
|
||||
</tr>
|
||||
</thead>
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td><a href="#axolotl.utils.models.ModelLoader.has_flash_attn">has_flash_attn</a></td>
|
||||
<td>Check if flash attention is installed</td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
</section>
|
||||
<section id="methods" class="level4">
|
||||
<h4 class="anchored" data-anchor-id="methods">Methods</h4>
|
||||
<table class="caption-top table">
|
||||
<thead>
|
||||
<tr class="header">
|
||||
<th>Name</th>
|
||||
<th>Description</th>
|
||||
</tr>
|
||||
</thead>
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td><a href="#axolotl.utils.models.ModelLoader.patch_llama_derived_model">patch_llama_derived_model</a></td>
|
||||
<td>Modify all llama derived models in one block</td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td><a href="#axolotl.utils.models.ModelLoader.patch_loss_llama">patch_loss_llama</a></td>
|
||||
<td>Patch loss functions and other optimizations</td>
|
||||
</tr>
|
||||
<tr class="odd">
|
||||
<td><a href="#axolotl.utils.models.ModelLoader.set_attention_config">set_attention_config</a></td>
|
||||
<td>sample packing uses custom FA2 patch</td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td><a href="#axolotl.utils.models.ModelLoader.set_auto_model_loader">set_auto_model_loader</a></td>
|
||||
<td>Set self.auto_model_loader. Defaults to <code>transformers.AutoModelForCausalLM</code></td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
<section id="axolotl.utils.models.ModelLoader.patch_llama_derived_model" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.utils.models.ModelLoader.patch_llama_derived_model">patch_llama_derived_model</h5>
|
||||
<div class="sourceCode" id="cb2"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb2-1"><a href="#cb2-1" aria-hidden="true" tabindex="-1"></a>utils.models.ModelLoader.patch_llama_derived_model()</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Modify all llama derived models in one block</p>
|
||||
</section>
|
||||
<section id="axolotl.utils.models.ModelLoader.patch_loss_llama" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.utils.models.ModelLoader.patch_loss_llama">patch_loss_llama</h5>
|
||||
<div class="sourceCode" id="cb3"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb3-1"><a href="#cb3-1" aria-hidden="true" tabindex="-1"></a>utils.models.ModelLoader.patch_loss_llama()</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Patch loss functions and other optimizations</p>
|
||||
</section>
|
||||
<section id="axolotl.utils.models.ModelLoader.set_attention_config" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.utils.models.ModelLoader.set_attention_config">set_attention_config</h5>
|
||||
<div class="sourceCode" id="cb4"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb4-1"><a href="#cb4-1" aria-hidden="true" tabindex="-1"></a>utils.models.ModelLoader.set_attention_config()</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>sample packing uses custom FA2 patch</p>
|
||||
</section>
|
||||
<section id="axolotl.utils.models.ModelLoader.set_auto_model_loader" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.utils.models.ModelLoader.set_auto_model_loader">set_auto_model_loader</h5>
|
||||
<div class="sourceCode" id="cb5"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb5-1"><a href="#cb5-1" aria-hidden="true" tabindex="-1"></a>utils.models.ModelLoader.set_auto_model_loader()</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Set self.auto_model_loader. Defaults to <code>transformers.AutoModelForCausalLM</code>
|
||||
(set at <code>__init__</code>). When using a multimodal model, <code>self.auto_model_loader</code>
|
||||
should be set according to the type of the model.</p>
|
||||
</section>
|
||||
</section>
|
||||
</section>
|
||||
</section>
|
||||
<section id="axolotl.loaders.tokenizer" class="level1">
|
||||
<h1>loaders.tokenizer</h1>
|
||||
<p><code>loaders.tokenizer</code></p>
|
||||
<p>Tokenizer loading functionality and associated utils</p>
|
||||
<section id="functions" class="level2">
|
||||
<h2 class="anchored" data-anchor-id="functions">Functions</h2>
|
||||
<table class="caption-top table">
|
||||
@@ -585,94 +478,33 @@ should be set according to the type of the model.</p>
|
||||
</thead>
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td><a href="#axolotl.utils.models.get_module_class_from_name">get_module_class_from_name</a></td>
|
||||
<td>Gets a class from a module by its name.</td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td><a href="#axolotl.utils.models.load_model">load_model</a></td>
|
||||
<td>Load a model for a given configuration and tokenizer.</td>
|
||||
</tr>
|
||||
<tr class="odd">
|
||||
<td><a href="#axolotl.utils.models.load_tokenizer">load_tokenizer</a></td>
|
||||
<td><a href="#axolotl.loaders.tokenizer.load_tokenizer">load_tokenizer</a></td>
|
||||
<td>Load and configure the tokenizer based on the provided config.</td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td><a href="#axolotl.utils.models.modify_tokenizer_files">modify_tokenizer_files</a></td>
|
||||
<td>Modify tokenizer files to replace added_tokens strings, save to output directory, and return the path to the modified tokenizer.</td>
|
||||
</tr>
|
||||
<tr class="odd">
|
||||
<td><a href="#axolotl.utils.models.setup_quantized_meta_for_peft">setup_quantized_meta_for_peft</a></td>
|
||||
<td>Replaces <code>quant_state.to</code> with a dummy function to prevent PEFT from moving <code>quant_state</code> to meta device</td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td><a href="#axolotl.utils.models.setup_quantized_peft_meta_for_training">setup_quantized_peft_meta_for_training</a></td>
|
||||
<td>Replaces dummy <code>quant_state.to</code> method with the original function to allow training to continue</td>
|
||||
<td><a href="#axolotl.loaders.tokenizer.modify_tokenizer_files">modify_tokenizer_files</a></td>
|
||||
<td>Modify tokenizer files to replace added_tokens strings, save to output directory,</td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
<section id="axolotl.utils.models.get_module_class_from_name" class="level3">
|
||||
<h3 class="anchored" data-anchor-id="axolotl.utils.models.get_module_class_from_name">get_module_class_from_name</h3>
|
||||
<div class="sourceCode" id="cb6"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb6-1"><a href="#cb6-1" aria-hidden="true" tabindex="-1"></a>utils.models.get_module_class_from_name(module, name)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Gets a class from a module by its name.</p>
|
||||
<section id="parameters" class="level4 doc-section doc-section-parameters">
|
||||
<h4 class="doc-section doc-section-parameters anchored" data-anchor-id="parameters">Parameters</h4>
|
||||
<table class="caption-top table">
|
||||
<colgroup>
|
||||
<col style="width: 10%">
|
||||
<col style="width: 25%">
|
||||
<col style="width: 47%">
|
||||
<col style="width: 16%">
|
||||
</colgroup>
|
||||
<thead>
|
||||
<tr class="header">
|
||||
<th>Name</th>
|
||||
<th>Type</th>
|
||||
<th>Description</th>
|
||||
<th>Default</th>
|
||||
</tr>
|
||||
</thead>
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td>module</td>
|
||||
<td><code>torch.nn.Module</code></td>
|
||||
<td>The module to get the class from.</td>
|
||||
<td><em>required</em></td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td>name</td>
|
||||
<td><code>str</code></td>
|
||||
<td>The name of the class.</td>
|
||||
<td><em>required</em></td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
</section>
|
||||
</section>
|
||||
<section id="axolotl.utils.models.load_model" class="level3">
|
||||
<h3 class="anchored" data-anchor-id="axolotl.utils.models.load_model">load_model</h3>
|
||||
<div class="sourceCode" id="cb7"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb7-1"><a href="#cb7-1" aria-hidden="true" tabindex="-1"></a>utils.models.load_model(</span>
|
||||
<span id="cb7-2"><a href="#cb7-2" aria-hidden="true" tabindex="-1"></a> cfg,</span>
|
||||
<span id="cb7-3"><a href="#cb7-3" aria-hidden="true" tabindex="-1"></a> tokenizer,</span>
|
||||
<span id="cb7-4"><a href="#cb7-4" aria-hidden="true" tabindex="-1"></a> <span class="op">*</span>,</span>
|
||||
<span id="cb7-5"><a href="#cb7-5" aria-hidden="true" tabindex="-1"></a> processor<span class="op">=</span><span class="va">None</span>,</span>
|
||||
<span id="cb7-6"><a href="#cb7-6" aria-hidden="true" tabindex="-1"></a> inference<span class="op">=</span><span class="va">False</span>,</span>
|
||||
<span id="cb7-7"><a href="#cb7-7" aria-hidden="true" tabindex="-1"></a> reference_model<span class="op">=</span><span class="va">False</span>,</span>
|
||||
<span id="cb7-8"><a href="#cb7-8" aria-hidden="true" tabindex="-1"></a> <span class="op">**</span>kwargs,</span>
|
||||
<span id="cb7-9"><a href="#cb7-9" aria-hidden="true" tabindex="-1"></a>)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Load a model for a given configuration and tokenizer.</p>
|
||||
</section>
|
||||
<section id="axolotl.utils.models.load_tokenizer" class="level3">
|
||||
<h3 class="anchored" data-anchor-id="axolotl.utils.models.load_tokenizer">load_tokenizer</h3>
|
||||
<div class="sourceCode" id="cb8"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb8-1"><a href="#cb8-1" aria-hidden="true" tabindex="-1"></a>utils.models.load_tokenizer(cfg)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<section id="axolotl.loaders.tokenizer.load_tokenizer" class="level3">
|
||||
<h3 class="anchored" data-anchor-id="axolotl.loaders.tokenizer.load_tokenizer">load_tokenizer</h3>
|
||||
<div class="sourceCode" id="cb1"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb1-1"><a href="#cb1-1" aria-hidden="true" tabindex="-1"></a>loaders.tokenizer.load_tokenizer(cfg)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Load and configure the tokenizer based on the provided config.</p>
|
||||
</section>
|
||||
<section id="axolotl.utils.models.modify_tokenizer_files" class="level3">
|
||||
<h3 class="anchored" data-anchor-id="axolotl.utils.models.modify_tokenizer_files">modify_tokenizer_files</h3>
|
||||
<div class="sourceCode" id="cb9"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb9-1"><a href="#cb9-1" aria-hidden="true" tabindex="-1"></a>utils.models.modify_tokenizer_files(tokenizer_path, token_mappings, output_dir)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Modify tokenizer files to replace added_tokens strings, save to output directory, and return the path to the modified tokenizer.</p>
|
||||
<p>This only works with reserved tokens that were added to the tokenizer, not tokens already part of the vocab.</p>
|
||||
<section id="parameters-1" class="level4 doc-section doc-section-parameters">
|
||||
<h4 class="doc-section doc-section-parameters anchored" data-anchor-id="parameters-1">Parameters</h4>
|
||||
<section id="axolotl.loaders.tokenizer.modify_tokenizer_files" class="level3">
|
||||
<h3 class="anchored" data-anchor-id="axolotl.loaders.tokenizer.modify_tokenizer_files">modify_tokenizer_files</h3>
|
||||
<div class="sourceCode" id="cb2"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb2-1"><a href="#cb2-1" aria-hidden="true" tabindex="-1"></a>loaders.tokenizer.modify_tokenizer_files(</span>
|
||||
<span id="cb2-2"><a href="#cb2-2" aria-hidden="true" tabindex="-1"></a> tokenizer_path,</span>
|
||||
<span id="cb2-3"><a href="#cb2-3" aria-hidden="true" tabindex="-1"></a> token_mappings,</span>
|
||||
<span id="cb2-4"><a href="#cb2-4" aria-hidden="true" tabindex="-1"></a> output_dir,</span>
|
||||
<span id="cb2-5"><a href="#cb2-5" aria-hidden="true" tabindex="-1"></a>)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Modify tokenizer files to replace added_tokens strings, save to output directory,
|
||||
and return the path to the modified tokenizer.</p>
|
||||
<p>This only works with reserved tokens that were added to the tokenizer, not tokens
|
||||
already part of the vocab.</p>
|
||||
<section id="parameters" class="level4 doc-section doc-section-parameters">
|
||||
<h4 class="doc-section doc-section-parameters anchored" data-anchor-id="parameters">Parameters</h4>
|
||||
<table class="caption-top table">
|
||||
<colgroup>
|
||||
<col style="width: 16%">
|
||||
@@ -697,7 +529,7 @@ should be set according to the type of the model.</p>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td>token_mappings</td>
|
||||
<td>Dict[int, str]</td>
|
||||
<td>dict[int, str]</td>
|
||||
<td>Dict mapping {token_id (int): new_token_string}</td>
|
||||
<td><em>required</em></td>
|
||||
</tr>
|
||||
@@ -729,22 +561,12 @@ should be set according to the type of the model.</p>
|
||||
</tbody>
|
||||
</table>
|
||||
<p>Ref: https://github.com/huggingface/transformers/issues/27974#issuecomment-1854188941</p>
|
||||
</section>
|
||||
</section>
|
||||
<section id="axolotl.utils.models.setup_quantized_meta_for_peft" class="level3">
|
||||
<h3 class="anchored" data-anchor-id="axolotl.utils.models.setup_quantized_meta_for_peft">setup_quantized_meta_for_peft</h3>
|
||||
<div class="sourceCode" id="cb10"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb10-1"><a href="#cb10-1" aria-hidden="true" tabindex="-1"></a>utils.models.setup_quantized_meta_for_peft(model)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Replaces <code>quant_state.to</code> with a dummy function to prevent PEFT from moving <code>quant_state</code> to meta device</p>
|
||||
</section>
|
||||
<section id="axolotl.utils.models.setup_quantized_peft_meta_for_training" class="level3">
|
||||
<h3 class="anchored" data-anchor-id="axolotl.utils.models.setup_quantized_peft_meta_for_training">setup_quantized_peft_meta_for_training</h3>
|
||||
<div class="sourceCode" id="cb11"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb11-1"><a href="#cb11-1" aria-hidden="true" tabindex="-1"></a>utils.models.setup_quantized_peft_meta_for_training(model)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Replaces dummy <code>quant_state.to</code> method with the original function to allow training to continue</p>
|
||||
|
||||
|
||||
</section>
|
||||
</section>
|
||||
</section>
|
||||
</section>
|
||||
|
||||
</main> <!-- /main -->
|
||||
<script id="quarto-html-after-body" type="application/javascript">
|
||||
@@ -7,7 +7,7 @@
|
||||
<meta name="viewport" content="width=device-width, initial-scale=1.0, user-scalable=yes">
|
||||
|
||||
|
||||
<title>utils.gradient_checkpointing.offload_cpu – Axolotl</title>
|
||||
<title>monkeypatch.gradient_checkpointing.offload_cpu – Axolotl</title>
|
||||
<style>
|
||||
code{white-space: pre-wrap;}
|
||||
span.smallcaps{font-variant: small-caps;}
|
||||
@@ -446,11 +446,11 @@ pre > code.sourceCode > span > a:first-child::before { text-decoration: underlin
|
||||
<h2 id="toc-title">On this page</h2>
|
||||
|
||||
<ul>
|
||||
<li><a href="#axolotl.utils.gradient_checkpointing.offload_cpu" id="toc-axolotl.utils.gradient_checkpointing.offload_cpu" class="nav-link active" data-scroll-target="#axolotl.utils.gradient_checkpointing.offload_cpu">utils.gradient_checkpointing.offload_cpu</a>
|
||||
<li><a href="#axolotl.monkeypatch.gradient_checkpointing.offload_cpu" id="toc-axolotl.monkeypatch.gradient_checkpointing.offload_cpu" class="nav-link active" data-scroll-target="#axolotl.monkeypatch.gradient_checkpointing.offload_cpu">monkeypatch.gradient_checkpointing.offload_cpu</a>
|
||||
<ul class="collapse">
|
||||
<li><a href="#classes" id="toc-classes" class="nav-link" data-scroll-target="#classes">Classes</a>
|
||||
<ul class="collapse">
|
||||
<li><a href="#axolotl.utils.gradient_checkpointing.offload_cpu.CPU_Offloaded_Gradient_Checkpointer" id="toc-axolotl.utils.gradient_checkpointing.offload_cpu.CPU_Offloaded_Gradient_Checkpointer" class="nav-link" data-scroll-target="#axolotl.utils.gradient_checkpointing.offload_cpu.CPU_Offloaded_Gradient_Checkpointer">CPU_Offloaded_Gradient_Checkpointer</a></li>
|
||||
<li><a href="#axolotl.monkeypatch.gradient_checkpointing.offload_cpu.CPU_Offloaded_Gradient_Checkpointer" id="toc-axolotl.monkeypatch.gradient_checkpointing.offload_cpu.CPU_Offloaded_Gradient_Checkpointer" class="nav-link" data-scroll-target="#axolotl.monkeypatch.gradient_checkpointing.offload_cpu.CPU_Offloaded_Gradient_Checkpointer">CPU_Offloaded_Gradient_Checkpointer</a></li>
|
||||
</ul></li>
|
||||
</ul></li>
|
||||
</ul>
|
||||
@@ -462,9 +462,9 @@ pre > code.sourceCode > span > a:first-child::before { text-decoration: underlin
|
||||
|
||||
|
||||
|
||||
<section id="axolotl.utils.gradient_checkpointing.offload_cpu" class="level1">
|
||||
<h1>utils.gradient_checkpointing.offload_cpu</h1>
|
||||
<p><code>utils.gradient_checkpointing.offload_cpu</code></p>
|
||||
<section id="axolotl.monkeypatch.gradient_checkpointing.offload_cpu" class="level1">
|
||||
<h1>monkeypatch.gradient_checkpointing.offload_cpu</h1>
|
||||
<p><code>monkeypatch.gradient_checkpointing.offload_cpu</code></p>
|
||||
<p>CPU offloaded checkpointing</p>
|
||||
<section id="classes" class="level2">
|
||||
<h2 class="anchored" data-anchor-id="classes">Classes</h2>
|
||||
@@ -477,14 +477,15 @@ pre > code.sourceCode > span > a:first-child::before { text-decoration: underlin
|
||||
</thead>
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td><a href="#axolotl.utils.gradient_checkpointing.offload_cpu.CPU_Offloaded_Gradient_Checkpointer">CPU_Offloaded_Gradient_Checkpointer</a></td>
|
||||
<td><a href="#axolotl.monkeypatch.gradient_checkpointing.offload_cpu.CPU_Offloaded_Gradient_Checkpointer">CPU_Offloaded_Gradient_Checkpointer</a></td>
|
||||
<td>Saves VRAM by smartly offloading to RAM.</td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
<section id="axolotl.utils.gradient_checkpointing.offload_cpu.CPU_Offloaded_Gradient_Checkpointer" class="level3">
|
||||
<h3 class="anchored" data-anchor-id="axolotl.utils.gradient_checkpointing.offload_cpu.CPU_Offloaded_Gradient_Checkpointer">CPU_Offloaded_Gradient_Checkpointer</h3>
|
||||
<div class="sourceCode" id="cb1"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb1-1"><a href="#cb1-1" aria-hidden="true" tabindex="-1"></a>utils.gradient_checkpointing.offload_cpu.CPU_Offloaded_Gradient_Checkpointer()</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<section id="axolotl.monkeypatch.gradient_checkpointing.offload_cpu.CPU_Offloaded_Gradient_Checkpointer" class="level3">
|
||||
<h3 class="anchored" data-anchor-id="axolotl.monkeypatch.gradient_checkpointing.offload_cpu.CPU_Offloaded_Gradient_Checkpointer">CPU_Offloaded_Gradient_Checkpointer</h3>
|
||||
<div class="sourceCode" id="cb1"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb1-1"><a href="#cb1-1" aria-hidden="true" tabindex="-1"></a>monkeypatch.gradient_checkpointing.offload_cpu.CPU_Offloaded_Gradient_Checkpointer(</span>
|
||||
<span id="cb1-2"><a href="#cb1-2" aria-hidden="true" tabindex="-1"></a>)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Saves VRAM by smartly offloading to RAM.
|
||||
Tiny hit to performance, since we mask the movement via non blocking calls.</p>
|
||||
|
||||
@@ -7,7 +7,7 @@
|
||||
<meta name="viewport" content="width=device-width, initial-scale=1.0, user-scalable=yes">
|
||||
|
||||
|
||||
<title>utils.gradient_checkpointing.offload_disk – Axolotl</title>
|
||||
<title>monkeypatch.gradient_checkpointing.offload_disk – Axolotl</title>
|
||||
<style>
|
||||
code{white-space: pre-wrap;}
|
||||
span.smallcaps{font-variant: small-caps;}
|
||||
@@ -446,12 +446,12 @@ pre > code.sourceCode > span > a:first-child::before { text-decoration: underlin
|
||||
<h2 id="toc-title">On this page</h2>
|
||||
|
||||
<ul>
|
||||
<li><a href="#axolotl.utils.gradient_checkpointing.offload_disk" id="toc-axolotl.utils.gradient_checkpointing.offload_disk" class="nav-link active" data-scroll-target="#axolotl.utils.gradient_checkpointing.offload_disk">utils.gradient_checkpointing.offload_disk</a>
|
||||
<li><a href="#axolotl.monkeypatch.gradient_checkpointing.offload_disk" id="toc-axolotl.monkeypatch.gradient_checkpointing.offload_disk" class="nav-link active" data-scroll-target="#axolotl.monkeypatch.gradient_checkpointing.offload_disk">monkeypatch.gradient_checkpointing.offload_disk</a>
|
||||
<ul class="collapse">
|
||||
<li><a href="#classes" id="toc-classes" class="nav-link" data-scroll-target="#classes">Classes</a>
|
||||
<ul class="collapse">
|
||||
<li><a href="#axolotl.utils.gradient_checkpointing.offload_disk.Disco" id="toc-axolotl.utils.gradient_checkpointing.offload_disk.Disco" class="nav-link" data-scroll-target="#axolotl.utils.gradient_checkpointing.offload_disk.Disco">Disco</a></li>
|
||||
<li><a href="#axolotl.utils.gradient_checkpointing.offload_disk.DiskOffloadManager" id="toc-axolotl.utils.gradient_checkpointing.offload_disk.DiskOffloadManager" class="nav-link" data-scroll-target="#axolotl.utils.gradient_checkpointing.offload_disk.DiskOffloadManager">DiskOffloadManager</a></li>
|
||||
<li><a href="#axolotl.monkeypatch.gradient_checkpointing.offload_disk.Disco" id="toc-axolotl.monkeypatch.gradient_checkpointing.offload_disk.Disco" class="nav-link" data-scroll-target="#axolotl.monkeypatch.gradient_checkpointing.offload_disk.Disco">Disco</a></li>
|
||||
<li><a href="#axolotl.monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager" id="toc-axolotl.monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager" class="nav-link" data-scroll-target="#axolotl.monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager">DiskOffloadManager</a></li>
|
||||
</ul></li>
|
||||
</ul></li>
|
||||
</ul>
|
||||
@@ -463,9 +463,9 @@ pre > code.sourceCode > span > a:first-child::before { text-decoration: underlin
|
||||
|
||||
|
||||
|
||||
<section id="axolotl.utils.gradient_checkpointing.offload_disk" class="level1">
|
||||
<h1>utils.gradient_checkpointing.offload_disk</h1>
|
||||
<p><code>utils.gradient_checkpointing.offload_disk</code></p>
|
||||
<section id="axolotl.monkeypatch.gradient_checkpointing.offload_disk" class="level1">
|
||||
<h1>monkeypatch.gradient_checkpointing.offload_disk</h1>
|
||||
<p><code>monkeypatch.gradient_checkpointing.offload_disk</code></p>
|
||||
<p>DISCO - DIsk-based Storage and Checkpointing with Optimized prefetching</p>
|
||||
<section id="classes" class="level2">
|
||||
<h2 class="anchored" data-anchor-id="classes">Classes</h2>
|
||||
@@ -478,18 +478,18 @@ pre > code.sourceCode > span > a:first-child::before { text-decoration: underlin
|
||||
</thead>
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td><a href="#axolotl.utils.gradient_checkpointing.offload_disk.Disco">Disco</a></td>
|
||||
<td><a href="#axolotl.monkeypatch.gradient_checkpointing.offload_disk.Disco">Disco</a></td>
|
||||
<td>Disco: DIsk-based Storage and Checkpointing with Optimized prefetching</td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td><a href="#axolotl.utils.gradient_checkpointing.offload_disk.DiskOffloadManager">DiskOffloadManager</a></td>
|
||||
<td><a href="#axolotl.monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager">DiskOffloadManager</a></td>
|
||||
<td>Manages offloaded tensors and handles prefetching in a separate thread.</td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
<section id="axolotl.utils.gradient_checkpointing.offload_disk.Disco" class="level3">
|
||||
<h3 class="anchored" data-anchor-id="axolotl.utils.gradient_checkpointing.offload_disk.Disco">Disco</h3>
|
||||
<div class="sourceCode" id="cb1"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb1-1"><a href="#cb1-1" aria-hidden="true" tabindex="-1"></a>utils.gradient_checkpointing.offload_disk.Disco()</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<section id="axolotl.monkeypatch.gradient_checkpointing.offload_disk.Disco" class="level3">
|
||||
<h3 class="anchored" data-anchor-id="axolotl.monkeypatch.gradient_checkpointing.offload_disk.Disco">Disco</h3>
|
||||
<div class="sourceCode" id="cb1"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb1-1"><a href="#cb1-1" aria-hidden="true" tabindex="-1"></a>monkeypatch.gradient_checkpointing.offload_disk.Disco()</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Disco: DIsk-based Storage and Checkpointing with Optimized prefetching
|
||||
Advanced disk-based gradient checkpointer with prefetching.</p>
|
||||
<section id="methods" class="level4">
|
||||
@@ -503,27 +503,30 @@ Advanced disk-based gradient checkpointer with prefetching.</p>
|
||||
</thead>
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td><a href="#axolotl.utils.gradient_checkpointing.offload_disk.Disco.backward">backward</a></td>
|
||||
<td><a href="#axolotl.monkeypatch.gradient_checkpointing.offload_disk.Disco.backward">backward</a></td>
|
||||
<td>Backward pass that loads activations from disk with prefetching</td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td><a href="#axolotl.utils.gradient_checkpointing.offload_disk.Disco.forward">forward</a></td>
|
||||
<td><a href="#axolotl.monkeypatch.gradient_checkpointing.offload_disk.Disco.forward">forward</a></td>
|
||||
<td>Forward pass that offloads activations to disk asynchronously</td>
|
||||
</tr>
|
||||
<tr class="odd">
|
||||
<td><a href="#axolotl.utils.gradient_checkpointing.offload_disk.Disco.get_instance">get_instance</a></td>
|
||||
<td><a href="#axolotl.monkeypatch.gradient_checkpointing.offload_disk.Disco.get_instance">get_instance</a></td>
|
||||
<td>Get or create the offload manager</td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
<section id="axolotl.utils.gradient_checkpointing.offload_disk.Disco.backward" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.utils.gradient_checkpointing.offload_disk.Disco.backward">backward</h5>
|
||||
<div class="sourceCode" id="cb2"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb2-1"><a href="#cb2-1" aria-hidden="true" tabindex="-1"></a>utils.gradient_checkpointing.offload_disk.Disco.backward(ctx, <span class="op">*</span>grad_outputs)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<section id="axolotl.monkeypatch.gradient_checkpointing.offload_disk.Disco.backward" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.monkeypatch.gradient_checkpointing.offload_disk.Disco.backward">backward</h5>
|
||||
<div class="sourceCode" id="cb2"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb2-1"><a href="#cb2-1" aria-hidden="true" tabindex="-1"></a>monkeypatch.gradient_checkpointing.offload_disk.Disco.backward(</span>
|
||||
<span id="cb2-2"><a href="#cb2-2" aria-hidden="true" tabindex="-1"></a> ctx,</span>
|
||||
<span id="cb2-3"><a href="#cb2-3" aria-hidden="true" tabindex="-1"></a> <span class="op">*</span>grad_outputs,</span>
|
||||
<span id="cb2-4"><a href="#cb2-4" aria-hidden="true" tabindex="-1"></a>)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Backward pass that loads activations from disk with prefetching</p>
|
||||
</section>
|
||||
<section id="axolotl.utils.gradient_checkpointing.offload_disk.Disco.forward" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.utils.gradient_checkpointing.offload_disk.Disco.forward">forward</h5>
|
||||
<div class="sourceCode" id="cb3"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb3-1"><a href="#cb3-1" aria-hidden="true" tabindex="-1"></a>utils.gradient_checkpointing.offload_disk.Disco.forward(</span>
|
||||
<section id="axolotl.monkeypatch.gradient_checkpointing.offload_disk.Disco.forward" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.monkeypatch.gradient_checkpointing.offload_disk.Disco.forward">forward</h5>
|
||||
<div class="sourceCode" id="cb3"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb3-1"><a href="#cb3-1" aria-hidden="true" tabindex="-1"></a>monkeypatch.gradient_checkpointing.offload_disk.Disco.forward(</span>
|
||||
<span id="cb3-2"><a href="#cb3-2" aria-hidden="true" tabindex="-1"></a> ctx,</span>
|
||||
<span id="cb3-3"><a href="#cb3-3" aria-hidden="true" tabindex="-1"></a> forward_function,</span>
|
||||
<span id="cb3-4"><a href="#cb3-4" aria-hidden="true" tabindex="-1"></a> hidden_states,</span>
|
||||
@@ -534,9 +537,9 @@ Advanced disk-based gradient checkpointer with prefetching.</p>
|
||||
<span id="cb3-9"><a href="#cb3-9" aria-hidden="true" tabindex="-1"></a>)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Forward pass that offloads activations to disk asynchronously</p>
|
||||
</section>
|
||||
<section id="axolotl.utils.gradient_checkpointing.offload_disk.Disco.get_instance" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.utils.gradient_checkpointing.offload_disk.Disco.get_instance">get_instance</h5>
|
||||
<div class="sourceCode" id="cb4"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb4-1"><a href="#cb4-1" aria-hidden="true" tabindex="-1"></a>utils.gradient_checkpointing.offload_disk.Disco.get_instance(</span>
|
||||
<section id="axolotl.monkeypatch.gradient_checkpointing.offload_disk.Disco.get_instance" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.monkeypatch.gradient_checkpointing.offload_disk.Disco.get_instance">get_instance</h5>
|
||||
<div class="sourceCode" id="cb4"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb4-1"><a href="#cb4-1" aria-hidden="true" tabindex="-1"></a>monkeypatch.gradient_checkpointing.offload_disk.Disco.get_instance(</span>
|
||||
<span id="cb4-2"><a href="#cb4-2" aria-hidden="true" tabindex="-1"></a> prefetch_size<span class="op">=</span><span class="dv">1</span>,</span>
|
||||
<span id="cb4-3"><a href="#cb4-3" aria-hidden="true" tabindex="-1"></a> prefetch_to_gpu<span class="op">=</span><span class="va">True</span>,</span>
|
||||
<span id="cb4-4"><a href="#cb4-4" aria-hidden="true" tabindex="-1"></a> save_workers<span class="op">=</span><span class="dv">4</span>,</span>
|
||||
@@ -545,9 +548,9 @@ Advanced disk-based gradient checkpointer with prefetching.</p>
|
||||
</section>
|
||||
</section>
|
||||
</section>
|
||||
<section id="axolotl.utils.gradient_checkpointing.offload_disk.DiskOffloadManager" class="level3">
|
||||
<h3 class="anchored" data-anchor-id="axolotl.utils.gradient_checkpointing.offload_disk.DiskOffloadManager">DiskOffloadManager</h3>
|
||||
<div class="sourceCode" id="cb5"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb5-1"><a href="#cb5-1" aria-hidden="true" tabindex="-1"></a>utils.gradient_checkpointing.offload_disk.DiskOffloadManager(</span>
|
||||
<section id="axolotl.monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager" class="level3">
|
||||
<h3 class="anchored" data-anchor-id="axolotl.monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager">DiskOffloadManager</h3>
|
||||
<div class="sourceCode" id="cb5"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb5-1"><a href="#cb5-1" aria-hidden="true" tabindex="-1"></a>monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager(</span>
|
||||
<span id="cb5-2"><a href="#cb5-2" aria-hidden="true" tabindex="-1"></a> <span class="va">self</span>,</span>
|
||||
<span id="cb5-3"><a href="#cb5-3" aria-hidden="true" tabindex="-1"></a> prefetch_size<span class="op">=</span><span class="dv">3</span>,</span>
|
||||
<span id="cb5-4"><a href="#cb5-4" aria-hidden="true" tabindex="-1"></a> prefetch_to_gpu<span class="op">=</span><span class="va">True</span>,</span>
|
||||
@@ -566,66 +569,68 @@ Includes synchronization to prevent race conditions.</p>
|
||||
</thead>
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td><a href="#axolotl.utils.gradient_checkpointing.offload_disk.DiskOffloadManager.cleanup">cleanup</a></td>
|
||||
<td><a href="#axolotl.monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager.cleanup">cleanup</a></td>
|
||||
<td>Clean up all temp files and stop prefetch thread with proper synchronization</td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td><a href="#axolotl.utils.gradient_checkpointing.offload_disk.DiskOffloadManager.cleanup_tensor">cleanup_tensor</a></td>
|
||||
<td><a href="#axolotl.monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager.cleanup_tensor">cleanup_tensor</a></td>
|
||||
<td>Clean up a specific tensor file after it’s been used</td>
|
||||
</tr>
|
||||
<tr class="odd">
|
||||
<td><a href="#axolotl.utils.gradient_checkpointing.offload_disk.DiskOffloadManager.load_tensor">load_tensor</a></td>
|
||||
<td><a href="#axolotl.monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager.load_tensor">load_tensor</a></td>
|
||||
<td>Load tensor from disk or prefetch cache with proper synchronization</td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td><a href="#axolotl.utils.gradient_checkpointing.offload_disk.DiskOffloadManager.save_tensor">save_tensor</a></td>
|
||||
<td><a href="#axolotl.monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager.save_tensor">save_tensor</a></td>
|
||||
<td>Save tensor to disk asynchronously and return file path with thread-safe operations</td>
|
||||
</tr>
|
||||
<tr class="odd">
|
||||
<td><a href="#axolotl.utils.gradient_checkpointing.offload_disk.DiskOffloadManager.trigger_prefetch">trigger_prefetch</a></td>
|
||||
<td><a href="#axolotl.monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager.trigger_prefetch">trigger_prefetch</a></td>
|
||||
<td>Trigger prefetching of the next N tensors with proper synchronization</td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td><a href="#axolotl.utils.gradient_checkpointing.offload_disk.DiskOffloadManager.wait_for_save">wait_for_save</a></td>
|
||||
<td><a href="#axolotl.monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager.wait_for_save">wait_for_save</a></td>
|
||||
<td>Wait for a tensor to be saved to disk</td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
<section id="axolotl.utils.gradient_checkpointing.offload_disk.DiskOffloadManager.cleanup" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.utils.gradient_checkpointing.offload_disk.DiskOffloadManager.cleanup">cleanup</h5>
|
||||
<div class="sourceCode" id="cb6"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb6-1"><a href="#cb6-1" aria-hidden="true" tabindex="-1"></a>utils.gradient_checkpointing.offload_disk.DiskOffloadManager.cleanup()</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<section id="axolotl.monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager.cleanup" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager.cleanup">cleanup</h5>
|
||||
<div class="sourceCode" id="cb6"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb6-1"><a href="#cb6-1" aria-hidden="true" tabindex="-1"></a>monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager.cleanup()</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Clean up all temp files and stop prefetch thread with proper synchronization</p>
|
||||
</section>
|
||||
<section id="axolotl.utils.gradient_checkpointing.offload_disk.DiskOffloadManager.cleanup_tensor" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.utils.gradient_checkpointing.offload_disk.DiskOffloadManager.cleanup_tensor">cleanup_tensor</h5>
|
||||
<div class="sourceCode" id="cb7"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb7-1"><a href="#cb7-1" aria-hidden="true" tabindex="-1"></a>utils.gradient_checkpointing.offload_disk.DiskOffloadManager.cleanup_tensor(</span>
|
||||
<section id="axolotl.monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager.cleanup_tensor" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager.cleanup_tensor">cleanup_tensor</h5>
|
||||
<div class="sourceCode" id="cb7"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb7-1"><a href="#cb7-1" aria-hidden="true" tabindex="-1"></a>monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager.cleanup_tensor(</span>
|
||||
<span id="cb7-2"><a href="#cb7-2" aria-hidden="true" tabindex="-1"></a> file_path,</span>
|
||||
<span id="cb7-3"><a href="#cb7-3" aria-hidden="true" tabindex="-1"></a>)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Clean up a specific tensor file after it’s been used</p>
|
||||
</section>
|
||||
<section id="axolotl.utils.gradient_checkpointing.offload_disk.DiskOffloadManager.load_tensor" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.utils.gradient_checkpointing.offload_disk.DiskOffloadManager.load_tensor">load_tensor</h5>
|
||||
<div class="sourceCode" id="cb8"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb8-1"><a href="#cb8-1" aria-hidden="true" tabindex="-1"></a>utils.gradient_checkpointing.offload_disk.DiskOffloadManager.load_tensor(</span>
|
||||
<section id="axolotl.monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager.load_tensor" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager.load_tensor">load_tensor</h5>
|
||||
<div class="sourceCode" id="cb8"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb8-1"><a href="#cb8-1" aria-hidden="true" tabindex="-1"></a>monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager.load_tensor(</span>
|
||||
<span id="cb8-2"><a href="#cb8-2" aria-hidden="true" tabindex="-1"></a> file_path,</span>
|
||||
<span id="cb8-3"><a href="#cb8-3" aria-hidden="true" tabindex="-1"></a> target_device<span class="op">=</span><span class="st">'cuda'</span>,</span>
|
||||
<span id="cb8-4"><a href="#cb8-4" aria-hidden="true" tabindex="-1"></a>)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Load tensor from disk or prefetch cache with proper synchronization</p>
|
||||
</section>
|
||||
<section id="axolotl.utils.gradient_checkpointing.offload_disk.DiskOffloadManager.save_tensor" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.utils.gradient_checkpointing.offload_disk.DiskOffloadManager.save_tensor">save_tensor</h5>
|
||||
<div class="sourceCode" id="cb9"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb9-1"><a href="#cb9-1" aria-hidden="true" tabindex="-1"></a>utils.gradient_checkpointing.offload_disk.DiskOffloadManager.save_tensor(tensor)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<section id="axolotl.monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager.save_tensor" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager.save_tensor">save_tensor</h5>
|
||||
<div class="sourceCode" id="cb9"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb9-1"><a href="#cb9-1" aria-hidden="true" tabindex="-1"></a>monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager.save_tensor(</span>
|
||||
<span id="cb9-2"><a href="#cb9-2" aria-hidden="true" tabindex="-1"></a> tensor,</span>
|
||||
<span id="cb9-3"><a href="#cb9-3" aria-hidden="true" tabindex="-1"></a>)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Save tensor to disk asynchronously and return file path with thread-safe operations</p>
|
||||
</section>
|
||||
<section id="axolotl.utils.gradient_checkpointing.offload_disk.DiskOffloadManager.trigger_prefetch" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.utils.gradient_checkpointing.offload_disk.DiskOffloadManager.trigger_prefetch">trigger_prefetch</h5>
|
||||
<div class="sourceCode" id="cb10"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb10-1"><a href="#cb10-1" aria-hidden="true" tabindex="-1"></a>utils.gradient_checkpointing.offload_disk.DiskOffloadManager.trigger_prefetch(</span>
|
||||
<section id="axolotl.monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager.trigger_prefetch" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager.trigger_prefetch">trigger_prefetch</h5>
|
||||
<div class="sourceCode" id="cb10"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb10-1"><a href="#cb10-1" aria-hidden="true" tabindex="-1"></a>monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager.trigger_prefetch(</span>
|
||||
<span id="cb10-2"><a href="#cb10-2" aria-hidden="true" tabindex="-1"></a> n<span class="op">=</span><span class="va">None</span>,</span>
|
||||
<span id="cb10-3"><a href="#cb10-3" aria-hidden="true" tabindex="-1"></a>)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Trigger prefetching of the next N tensors with proper synchronization</p>
|
||||
</section>
|
||||
<section id="axolotl.utils.gradient_checkpointing.offload_disk.DiskOffloadManager.wait_for_save" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.utils.gradient_checkpointing.offload_disk.DiskOffloadManager.wait_for_save">wait_for_save</h5>
|
||||
<div class="sourceCode" id="cb11"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb11-1"><a href="#cb11-1" aria-hidden="true" tabindex="-1"></a>utils.gradient_checkpointing.offload_disk.DiskOffloadManager.wait_for_save(</span>
|
||||
<section id="axolotl.monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager.wait_for_save" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager.wait_for_save">wait_for_save</h5>
|
||||
<div class="sourceCode" id="cb11"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb11-1"><a href="#cb11-1" aria-hidden="true" tabindex="-1"></a>monkeypatch.gradient_checkpointing.offload_disk.DiskOffloadManager.wait_for_save(</span>
|
||||
<span id="cb11-2"><a href="#cb11-2" aria-hidden="true" tabindex="-1"></a> file_path,</span>
|
||||
<span id="cb11-3"><a href="#cb11-3" aria-hidden="true" tabindex="-1"></a> timeout<span class="op">=</span><span class="va">None</span>,</span>
|
||||
<span id="cb11-4"><a href="#cb11-4" aria-hidden="true" tabindex="-1"></a>)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
@@ -525,19 +525,19 @@ sequential packing (preserving original sequence order).</p>
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td><a href="#axolotl.utils.samplers.multipack.MultipackBatchSampler.efficiency">efficiency</a></td>
|
||||
<td>Calculate the packing efficiency (ratio of tokens used to total token slots)</td>
|
||||
<td>Calculate the packing efficiency (ratio of tokens used to total token slots).</td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td><a href="#axolotl.utils.samplers.multipack.MultipackBatchSampler.gather_efficiency">gather_efficiency</a></td>
|
||||
<td>Gather and synchronize packing efficiency estimates across all distributed ranks</td>
|
||||
<td>Gather and synchronize packing efficiency estimates across all distributed</td>
|
||||
</tr>
|
||||
<tr class="odd">
|
||||
<td><a href="#axolotl.utils.samplers.multipack.MultipackBatchSampler.gather_len_batches">gather_len_batches</a></td>
|
||||
<td>Gather and synchronize batch counts across all distributed ranks</td>
|
||||
<td>Gather and synchronize batch counts across all distributed ranks. Returns</td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td><a href="#axolotl.utils.samplers.multipack.MultipackBatchSampler.generate_batches">generate_batches</a></td>
|
||||
<td>Generate packed batches for training</td>
|
||||
<td>Generate packed batches for training.</td>
|
||||
</tr>
|
||||
<tr class="odd">
|
||||
<td><a href="#axolotl.utils.samplers.multipack.MultipackBatchSampler.set_epoch">set_epoch</a></td>
|
||||
@@ -548,32 +548,56 @@ sequential packing (preserving original sequence order).</p>
|
||||
<section id="axolotl.utils.samplers.multipack.MultipackBatchSampler.efficiency" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.utils.samplers.multipack.MultipackBatchSampler.efficiency">efficiency</h5>
|
||||
<div class="sourceCode" id="cb2"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb2-1"><a href="#cb2-1" aria-hidden="true" tabindex="-1"></a>utils.samplers.multipack.MultipackBatchSampler.efficiency()</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Calculate the packing efficiency (ratio of tokens used to total token slots)
|
||||
Higher is better - 1.0 would mean perfect packing with no wasted space</p>
|
||||
<p>Calculate the packing efficiency (ratio of tokens used to total token slots).
|
||||
Higher is better - 1.0 would mean perfect packing with no wasted space.</p>
|
||||
</section>
|
||||
<section id="axolotl.utils.samplers.multipack.MultipackBatchSampler.gather_efficiency" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.utils.samplers.multipack.MultipackBatchSampler.gather_efficiency">gather_efficiency</h5>
|
||||
<div class="sourceCode" id="cb3"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb3-1"><a href="#cb3-1" aria-hidden="true" tabindex="-1"></a>utils.samplers.multipack.MultipackBatchSampler.gather_efficiency()</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Gather and synchronize packing efficiency estimates across all distributed ranks
|
||||
Returns a conservative efficiency estimate based on the measurements</p>
|
||||
<p>Gather and synchronize packing efficiency estimates across all distributed
|
||||
ranks.</p>
|
||||
<section id="returns" class="level6 doc-section doc-section-returns">
|
||||
<h6 class="doc-section doc-section-returns anchored" data-anchor-id="returns">Returns</h6>
|
||||
<table class="caption-top table">
|
||||
<colgroup>
|
||||
<col style="width: 10%">
|
||||
<col style="width: 10%">
|
||||
<col style="width: 79%">
|
||||
</colgroup>
|
||||
<thead>
|
||||
<tr class="header">
|
||||
<th>Name</th>
|
||||
<th>Type</th>
|
||||
<th>Description</th>
|
||||
</tr>
|
||||
</thead>
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td></td>
|
||||
<td>float</td>
|
||||
<td>A conservative efficiency estimate based on the measurements.</td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
</section>
|
||||
</section>
|
||||
<section id="axolotl.utils.samplers.multipack.MultipackBatchSampler.gather_len_batches" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.utils.samplers.multipack.MultipackBatchSampler.gather_len_batches">gather_len_batches</h5>
|
||||
<div class="sourceCode" id="cb4"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb4-1"><a href="#cb4-1" aria-hidden="true" tabindex="-1"></a>utils.samplers.multipack.MultipackBatchSampler.gather_len_batches(num)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Gather and synchronize batch counts across all distributed ranks
|
||||
Returns the minimum number of batches available on any rank</p>
|
||||
<p>Gather and synchronize batch counts across all distributed ranks. Returns
|
||||
the minimum number of batches available on any rank.</p>
|
||||
</section>
|
||||
<section id="axolotl.utils.samplers.multipack.MultipackBatchSampler.generate_batches" class="level5">
|
||||
<h5 class="anchored" data-anchor-id="axolotl.utils.samplers.multipack.MultipackBatchSampler.generate_batches">generate_batches</h5>
|
||||
<div class="sourceCode" id="cb5"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb5-1"><a href="#cb5-1" aria-hidden="true" tabindex="-1"></a>utils.samplers.multipack.MultipackBatchSampler.generate_batches(set_stats<span class="op">=</span><span class="va">False</span>)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Generate packed batches for training</p>
|
||||
<p>Generate packed batches for training.</p>
|
||||
<section id="parameters" class="level6 doc-section doc-section-parameters">
|
||||
<h6 class="doc-section doc-section-parameters anchored" data-anchor-id="parameters">Parameters</h6>
|
||||
<table class="caption-top table">
|
||||
<colgroup>
|
||||
<col style="width: 15%">
|
||||
<col style="width: 11%">
|
||||
<col style="width: 57%">
|
||||
<col style="width: 58%">
|
||||
<col style="width: 15%">
|
||||
</colgroup>
|
||||
<thead>
|
||||
@@ -587,20 +611,20 @@ Returns the minimum number of batches available on any rank</p>
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td>set_stats</td>
|
||||
<td></td>
|
||||
<td>Whether to update efficiency statistics</td>
|
||||
<td>bool</td>
|
||||
<td>Whether to update efficiency statistics.</td>
|
||||
<td><code>False</code></td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
</section>
|
||||
<section id="returns" class="level6 doc-section doc-section-returns">
|
||||
<h6 class="doc-section doc-section-returns anchored" data-anchor-id="returns">Returns</h6>
|
||||
<section id="returns-1" class="level6 doc-section doc-section-returns">
|
||||
<h6 class="doc-section doc-section-returns anchored" data-anchor-id="returns-1">Returns</h6>
|
||||
<table class="caption-top table">
|
||||
<colgroup>
|
||||
<col style="width: 10%">
|
||||
<col style="width: 10%">
|
||||
<col style="width: 78%">
|
||||
<col style="width: 5%">
|
||||
<col style="width: 20%">
|
||||
<col style="width: 74%">
|
||||
</colgroup>
|
||||
<thead>
|
||||
<tr class="header">
|
||||
@@ -612,13 +636,8 @@ Returns the minimum number of batches available on any rank</p>
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td></td>
|
||||
<td></td>
|
||||
<td>List of batches, where each batch contains multiple bins,</td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td></td>
|
||||
<td></td>
|
||||
<td>and each bin contains multiple sequence indices</td>
|
||||
<td>list[list[list[int]]]</td>
|
||||
<td>List of batches, where each batch contains multiple bins, and each bin contains multiple sequence indices.</td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
@@ -644,19 +663,19 @@ Returns the minimum number of batches available on any rank</p>
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td><a href="#axolotl.utils.samplers.multipack.allocate_sequentially">allocate_sequentially</a></td>
|
||||
<td>Sequential allocator that preserves example order</td>
|
||||
<td>Sequential allocator that preserves example order.</td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td><a href="#axolotl.utils.samplers.multipack.ffd_check">ffd_check</a></td>
|
||||
<td>First-fit-decreasing bin packing algorithm check</td>
|
||||
<td>First-fit-decreasing bin packing algorithm check.</td>
|
||||
</tr>
|
||||
<tr class="odd">
|
||||
<td><a href="#axolotl.utils.samplers.multipack.pack_group">pack_group</a></td>
|
||||
<td>Pack a group of sequences into bins using First-Fit Decreasing algorithm</td>
|
||||
<td>Pack a group of sequences into bins using First-Fit Decreasing algorithm.</td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td><a href="#axolotl.utils.samplers.multipack.pack_parallel">pack_parallel</a></td>
|
||||
<td>Pack sequences into bins using parallel processing</td>
|
||||
<td>Pack sequences into bins using parallel processing.</td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
@@ -668,11 +687,172 @@ Returns the minimum number of batches available on any rank</p>
|
||||
<span id="cb7-4"><a href="#cb7-4" aria-hidden="true" tabindex="-1"></a> bin_capacity,</span>
|
||||
<span id="cb7-5"><a href="#cb7-5" aria-hidden="true" tabindex="-1"></a> num_ranks,</span>
|
||||
<span id="cb7-6"><a href="#cb7-6" aria-hidden="true" tabindex="-1"></a>)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Sequential allocator that preserves example order</p>
|
||||
<p>Sequential allocator that preserves example order.</p>
|
||||
<section id="parameters-1" class="level4 doc-section doc-section-parameters">
|
||||
<h4 class="doc-section doc-section-parameters anchored" data-anchor-id="parameters-1">Parameters</h4>
|
||||
<table class="caption-top table">
|
||||
<colgroup>
|
||||
<col style="width: 18%">
|
||||
<col style="width: 12%">
|
||||
<col style="width: 55%">
|
||||
<col style="width: 12%">
|
||||
</colgroup>
|
||||
<thead>
|
||||
<tr class="header">
|
||||
<th>Name</th>
|
||||
<th>Type</th>
|
||||
<th>Description</th>
|
||||
<th>Default</th>
|
||||
</tr>
|
||||
</thead>
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td>sequence_lengths</td>
|
||||
<td>np.ndarray</td>
|
||||
<td>The lengths of all examples.</td>
|
||||
<td><em>required</em></td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td>rank</td>
|
||||
<td>int</td>
|
||||
<td>The current rank (for distributed training).</td>
|
||||
<td><em>required</em></td>
|
||||
</tr>
|
||||
<tr class="odd">
|
||||
<td>bin_capacity</td>
|
||||
<td>int</td>
|
||||
<td>The capacity of each bin (maximum sequence length).</td>
|
||||
<td><em>required</em></td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td>num_ranks</td>
|
||||
<td>int</td>
|
||||
<td>Number of ranks (processes / GPUs).</td>
|
||||
<td><em>required</em></td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
</section>
|
||||
<section id="returns-2" class="level4 doc-section doc-section-returns">
|
||||
<h4 class="doc-section doc-section-returns anchored" data-anchor-id="returns-2">Returns</h4>
|
||||
<table class="caption-top table">
|
||||
<colgroup>
|
||||
<col style="width: 15%">
|
||||
<col style="width: 17%">
|
||||
<col style="width: 66%">
|
||||
</colgroup>
|
||||
<thead>
|
||||
<tr class="header">
|
||||
<th>Name</th>
|
||||
<th>Type</th>
|
||||
<th>Description</th>
|
||||
</tr>
|
||||
</thead>
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td>rank_batches</td>
|
||||
<td>list[list[int]]</td>
|
||||
<td>List of batches for the current rank.</td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td>total_tokens_used</td>
|
||||
<td>int</td>
|
||||
<td>Number of actual example tokens.</td>
|
||||
</tr>
|
||||
<tr class="odd">
|
||||
<td>total_token_slots</td>
|
||||
<td>int</td>
|
||||
<td>Maximum theoretical number of example tokens (number of bins * bin capacity).</td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
</section>
|
||||
</section>
|
||||
<section id="axolotl.utils.samplers.multipack.ffd_check" class="level3">
|
||||
<h3 class="anchored" data-anchor-id="axolotl.utils.samplers.multipack.ffd_check">ffd_check</h3>
|
||||
<div class="sourceCode" id="cb8"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb8-1"><a href="#cb8-1" aria-hidden="true" tabindex="-1"></a>utils.samplers.multipack.ffd_check(sequence_lengths, bin_capacity, num_bins)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>First-fit-decreasing bin packing algorithm check.</p>
|
||||
<p>Checks if sequences with the given lengths could fit in the specified number of
|
||||
bins.</p>
|
||||
<section id="parameters-2" class="level4 doc-section doc-section-parameters">
|
||||
<h4 class="doc-section doc-section-parameters anchored" data-anchor-id="parameters-2">Parameters</h4>
|
||||
<table class="caption-top table">
|
||||
<colgroup>
|
||||
<col style="width: 24%">
|
||||
<col style="width: 16%">
|
||||
<col style="width: 42%">
|
||||
<col style="width: 16%">
|
||||
</colgroup>
|
||||
<thead>
|
||||
<tr class="header">
|
||||
<th>Name</th>
|
||||
<th>Type</th>
|
||||
<th>Description</th>
|
||||
<th>Default</th>
|
||||
</tr>
|
||||
</thead>
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td>sequence_lengths</td>
|
||||
<td>np.ndarray</td>
|
||||
<td>Array of sequence lengths.</td>
|
||||
<td><em>required</em></td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td>bin_capacity</td>
|
||||
<td>int</td>
|
||||
<td>Maximum capacity of each bin.</td>
|
||||
<td><em>required</em></td>
|
||||
</tr>
|
||||
<tr class="odd">
|
||||
<td>num_bins</td>
|
||||
<td>int</td>
|
||||
<td>Number of bins available.</td>
|
||||
<td><em>required</em></td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
</section>
|
||||
<section id="returns-3" class="level4 doc-section doc-section-returns">
|
||||
<h4 class="doc-section doc-section-returns anchored" data-anchor-id="returns-3">Returns</h4>
|
||||
<table class="caption-top table">
|
||||
<colgroup>
|
||||
<col style="width: 10%">
|
||||
<col style="width: 10%">
|
||||
<col style="width: 78%">
|
||||
</colgroup>
|
||||
<thead>
|
||||
<tr class="header">
|
||||
<th>Name</th>
|
||||
<th>Type</th>
|
||||
<th>Description</th>
|
||||
</tr>
|
||||
</thead>
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td></td>
|
||||
<td>bool</td>
|
||||
<td><code>True</code> if all sequences can be packed, <code>False</code> otherwise.</td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
</section>
|
||||
</section>
|
||||
<section id="axolotl.utils.samplers.multipack.pack_group" class="level3">
|
||||
<h3 class="anchored" data-anchor-id="axolotl.utils.samplers.multipack.pack_group">pack_group</h3>
|
||||
<div class="sourceCode" id="cb9"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb9-1"><a href="#cb9-1" aria-hidden="true" tabindex="-1"></a>utils.samplers.multipack.pack_group(</span>
|
||||
<span id="cb9-2"><a href="#cb9-2" aria-hidden="true" tabindex="-1"></a> sequence_lengths,</span>
|
||||
<span id="cb9-3"><a href="#cb9-3" aria-hidden="true" tabindex="-1"></a> group_offset,</span>
|
||||
<span id="cb9-4"><a href="#cb9-4" aria-hidden="true" tabindex="-1"></a> bin_capacity,</span>
|
||||
<span id="cb9-5"><a href="#cb9-5" aria-hidden="true" tabindex="-1"></a> max_bins,</span>
|
||||
<span id="cb9-6"><a href="#cb9-6" aria-hidden="true" tabindex="-1"></a> bin_size,</span>
|
||||
<span id="cb9-7"><a href="#cb9-7" aria-hidden="true" tabindex="-1"></a> safe_mode<span class="op">=</span><span class="va">True</span>,</span>
|
||||
<span id="cb9-8"><a href="#cb9-8" aria-hidden="true" tabindex="-1"></a>)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Pack a group of sequences into bins using First-Fit Decreasing algorithm.</p>
|
||||
<section id="parameters-3" class="level4 doc-section doc-section-parameters">
|
||||
<h4 class="doc-section doc-section-parameters anchored" data-anchor-id="parameters-3">Parameters</h4>
|
||||
<table class="caption-top table">
|
||||
<colgroup>
|
||||
<col style="width: 19%">
|
||||
<col style="width: 12%">
|
||||
<col style="width: 55%">
|
||||
@@ -690,209 +870,49 @@ Returns the minimum number of batches available on any rank</p>
|
||||
<tr class="odd">
|
||||
<td>sequence_lengths</td>
|
||||
<td>np.ndarray</td>
|
||||
<td>The lengths of all examples</td>
|
||||
<td><em>required</em></td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td>rank</td>
|
||||
<td>int</td>
|
||||
<td>The current rank (for distributed training)</td>
|
||||
<td><em>required</em></td>
|
||||
</tr>
|
||||
<tr class="odd">
|
||||
<td>bin_capacity</td>
|
||||
<td>int</td>
|
||||
<td>The capacity of each bin (maximum sequence length)</td>
|
||||
<td><em>required</em></td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td>num_ranks</td>
|
||||
<td>int</td>
|
||||
<td>Number of ranks (processes/GPUs)</td>
|
||||
<td><em>required</em></td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
</section>
|
||||
<section id="returns-1" class="level4 doc-section doc-section-returns">
|
||||
<h4 class="doc-section doc-section-returns anchored" data-anchor-id="returns-1">Returns</h4>
|
||||
<table class="caption-top table">
|
||||
<colgroup>
|
||||
<col style="width: 18%">
|
||||
<col style="width: 7%">
|
||||
<col style="width: 74%">
|
||||
</colgroup>
|
||||
<thead>
|
||||
<tr class="header">
|
||||
<th>Name</th>
|
||||
<th>Type</th>
|
||||
<th>Description</th>
|
||||
</tr>
|
||||
</thead>
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td>rank_batches</td>
|
||||
<td></td>
|
||||
<td>List of batches for the current rank</td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td>total_tokens_used</td>
|
||||
<td></td>
|
||||
<td>Number of actual example tokens</td>
|
||||
</tr>
|
||||
<tr class="odd">
|
||||
<td>total_token_slots</td>
|
||||
<td></td>
|
||||
<td>Maximum theoretical number of example tokens (number of bins * bin capacity)</td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
</section>
|
||||
</section>
|
||||
<section id="axolotl.utils.samplers.multipack.ffd_check" class="level3">
|
||||
<h3 class="anchored" data-anchor-id="axolotl.utils.samplers.multipack.ffd_check">ffd_check</h3>
|
||||
<div class="sourceCode" id="cb8"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb8-1"><a href="#cb8-1" aria-hidden="true" tabindex="-1"></a>utils.samplers.multipack.ffd_check(sequence_lengths, bin_capacity, num_bins)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>First-fit-decreasing bin packing algorithm check</p>
|
||||
<p>Checks if sequences with the given lengths could fit in the specified number of bins</p>
|
||||
<section id="parameters-2" class="level4 doc-section doc-section-parameters">
|
||||
<h4 class="doc-section doc-section-parameters anchored" data-anchor-id="parameters-2">Parameters</h4>
|
||||
<table class="caption-top table">
|
||||
<colgroup>
|
||||
<col style="width: 25%">
|
||||
<col style="width: 16%">
|
||||
<col style="width: 41%">
|
||||
<col style="width: 16%">
|
||||
</colgroup>
|
||||
<thead>
|
||||
<tr class="header">
|
||||
<th>Name</th>
|
||||
<th>Type</th>
|
||||
<th>Description</th>
|
||||
<th>Default</th>
|
||||
</tr>
|
||||
</thead>
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td>sequence_lengths</td>
|
||||
<td>np.ndarray</td>
|
||||
<td>Array of sequence lengths</td>
|
||||
<td><em>required</em></td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td>bin_capacity</td>
|
||||
<td>int</td>
|
||||
<td>Maximum capacity of each bin</td>
|
||||
<td><em>required</em></td>
|
||||
</tr>
|
||||
<tr class="odd">
|
||||
<td>num_bins</td>
|
||||
<td>int</td>
|
||||
<td>Number of bins available</td>
|
||||
<td><em>required</em></td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
</section>
|
||||
<section id="returns-2" class="level4 doc-section doc-section-returns">
|
||||
<h4 class="doc-section doc-section-returns anchored" data-anchor-id="returns-2">Returns</h4>
|
||||
<table class="caption-top table">
|
||||
<colgroup>
|
||||
<col style="width: 11%">
|
||||
<col style="width: 11%">
|
||||
<col style="width: 77%">
|
||||
</colgroup>
|
||||
<thead>
|
||||
<tr class="header">
|
||||
<th>Name</th>
|
||||
<th>Type</th>
|
||||
<th>Description</th>
|
||||
</tr>
|
||||
</thead>
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td></td>
|
||||
<td></td>
|
||||
<td>True if all sequences can be packed, False otherwise</td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
</section>
|
||||
</section>
|
||||
<section id="axolotl.utils.samplers.multipack.pack_group" class="level3">
|
||||
<h3 class="anchored" data-anchor-id="axolotl.utils.samplers.multipack.pack_group">pack_group</h3>
|
||||
<div class="sourceCode" id="cb9"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb9-1"><a href="#cb9-1" aria-hidden="true" tabindex="-1"></a>utils.samplers.multipack.pack_group(</span>
|
||||
<span id="cb9-2"><a href="#cb9-2" aria-hidden="true" tabindex="-1"></a> sequence_lengths,</span>
|
||||
<span id="cb9-3"><a href="#cb9-3" aria-hidden="true" tabindex="-1"></a> group_offset,</span>
|
||||
<span id="cb9-4"><a href="#cb9-4" aria-hidden="true" tabindex="-1"></a> bin_capacity,</span>
|
||||
<span id="cb9-5"><a href="#cb9-5" aria-hidden="true" tabindex="-1"></a> max_bins,</span>
|
||||
<span id="cb9-6"><a href="#cb9-6" aria-hidden="true" tabindex="-1"></a> bin_size,</span>
|
||||
<span id="cb9-7"><a href="#cb9-7" aria-hidden="true" tabindex="-1"></a> safe_mode<span class="op">=</span><span class="va">True</span>,</span>
|
||||
<span id="cb9-8"><a href="#cb9-8" aria-hidden="true" tabindex="-1"></a>)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Pack a group of sequences into bins using First-Fit Decreasing algorithm</p>
|
||||
<section id="parameters-3" class="level4 doc-section doc-section-parameters">
|
||||
<h4 class="doc-section doc-section-parameters anchored" data-anchor-id="parameters-3">Parameters</h4>
|
||||
<table class="caption-top table">
|
||||
<colgroup>
|
||||
<col style="width: 19%">
|
||||
<col style="width: 12%">
|
||||
<col style="width: 54%">
|
||||
<col style="width: 12%">
|
||||
</colgroup>
|
||||
<thead>
|
||||
<tr class="header">
|
||||
<th>Name</th>
|
||||
<th>Type</th>
|
||||
<th>Description</th>
|
||||
<th>Default</th>
|
||||
</tr>
|
||||
</thead>
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td>sequence_lengths</td>
|
||||
<td>np.ndarray</td>
|
||||
<td>Array of sequence lengths</td>
|
||||
<td>Array of sequence lengths.</td>
|
||||
<td><em>required</em></td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td>group_offset</td>
|
||||
<td>int</td>
|
||||
<td>Offset to apply to indices when returning results</td>
|
||||
<td>Offset to apply to indices when returning results.</td>
|
||||
<td><em>required</em></td>
|
||||
</tr>
|
||||
<tr class="odd">
|
||||
<td>bin_capacity</td>
|
||||
<td>int</td>
|
||||
<td>Maximum capacity of each bin</td>
|
||||
<td>Maximum capacity of each bin.</td>
|
||||
<td><em>required</em></td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td>max_bins</td>
|
||||
<td>int</td>
|
||||
<td>Maximum number of bins to use</td>
|
||||
<td>Maximum number of bins to use.</td>
|
||||
<td><em>required</em></td>
|
||||
</tr>
|
||||
<tr class="odd">
|
||||
<td>bin_size</td>
|
||||
<td>int</td>
|
||||
<td>Maximum number of sequences per bin</td>
|
||||
<td>Maximum number of sequences per bin.</td>
|
||||
<td><em>required</em></td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td>safe_mode</td>
|
||||
<td>bool</td>
|
||||
<td>If True, use a more conservative packing approach</td>
|
||||
<td>If True, use a more conservative packing approach.</td>
|
||||
<td><code>True</code></td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
</section>
|
||||
<section id="returns-3" class="level4 doc-section doc-section-returns">
|
||||
<h4 class="doc-section doc-section-returns anchored" data-anchor-id="returns-3">Returns</h4>
|
||||
<section id="returns-4" class="level4 doc-section doc-section-returns">
|
||||
<h4 class="doc-section doc-section-returns anchored" data-anchor-id="returns-4">Returns</h4>
|
||||
<table class="caption-top table">
|
||||
<colgroup>
|
||||
<col style="width: 8%">
|
||||
<col style="width: 8%">
|
||||
<col style="width: 82%">
|
||||
<col style="width: 7%">
|
||||
<col style="width: 20%">
|
||||
<col style="width: 72%">
|
||||
</colgroup>
|
||||
<thead>
|
||||
<tr class="header">
|
||||
@@ -904,8 +924,8 @@ Returns the minimum number of batches available on any rank</p>
|
||||
<tbody>
|
||||
<tr class="odd">
|
||||
<td></td>
|
||||
<td></td>
|
||||
<td>List of bins, where each bin contains indices of sequences assigned to it</td>
|
||||
<td>list[list[int]]</td>
|
||||
<td>List of bins, where each bin contains indices of sequences assigned to it.</td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
@@ -922,7 +942,7 @@ Returns the minimum number of batches available on any rank</p>
|
||||
<span id="cb10-7"><a href="#cb10-7" aria-hidden="true" tabindex="-1"></a> safe_mode<span class="op">=</span><span class="va">True</span>,</span>
|
||||
<span id="cb10-8"><a href="#cb10-8" aria-hidden="true" tabindex="-1"></a> mp_start_method<span class="op">=</span><span class="st">'spawn'</span>,</span>
|
||||
<span id="cb10-9"><a href="#cb10-9" aria-hidden="true" tabindex="-1"></a>)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
|
||||
<p>Pack sequences into bins using parallel processing</p>
|
||||
<p>Pack sequences into bins using parallel processing.</p>
|
||||
<section id="parameters-4" class="level4 doc-section doc-section-parameters">
|
||||
<h4 class="doc-section doc-section-parameters anchored" data-anchor-id="parameters-4">Parameters</h4>
|
||||
<table class="caption-top table">
|
||||
@@ -944,37 +964,37 @@ Returns the minimum number of batches available on any rank</p>
|
||||
<tr class="odd">
|
||||
<td>sequence_lengths</td>
|
||||
<td>np.ndarray</td>
|
||||
<td>Array of sequence lengths</td>
|
||||
<td>Array of sequence lengths.</td>
|
||||
<td><em>required</em></td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td>bin_capacity</td>
|
||||
<td>int</td>
|
||||
<td>Maximum capacity of each bin as total number of tokens</td>
|
||||
<td>Maximum capacity of each bin as total number of tokens.</td>
|
||||
<td><em>required</em></td>
|
||||
</tr>
|
||||
<tr class="odd">
|
||||
<td>group_size</td>
|
||||
<td>int</td>
|
||||
<td>Number of sequences to process in each group</td>
|
||||
<td>Number of sequences to process in each group.</td>
|
||||
<td><em>required</em></td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td>bin_size</td>
|
||||
<td>int</td>
|
||||
<td>Maximum number of bins to use</td>
|
||||
<td>Maximum number of bins to use.</td>
|
||||
<td><em>required</em></td>
|
||||
</tr>
|
||||
<tr class="odd">
|
||||
<td>num_processes</td>
|
||||
<td>int | None</td>
|
||||
<td>Number of parallel processes to use</td>
|
||||
<td>Number of parallel processes to use.</td>
|
||||
<td><code>None</code></td>
|
||||
</tr>
|
||||
<tr class="even">
|
||||
<td>safe_mode</td>
|
||||
<td>bool</td>
|
||||
<td>If True, use a more conservative packing approach</td>
|
||||
<td>If True, use a more conservative packing approach.</td>
|
||||
<td><code>True</code></td>
|
||||
</tr>
|
||||
<tr class="odd">
|
||||
@@ -986,7 +1006,7 @@ Returns the minimum number of batches available on any rank</p>
|
||||
</tbody>
|
||||
</table>
|
||||
<p>Returns:
|
||||
List of bins, where each bin contains indices of sequences assigned to it</p>
|
||||
List of bins, where each bin contains indices of sequences assigned to it.</p>
|
||||
|
||||
|
||||
</section>
|
||||
|
||||
@@ -641,7 +641,7 @@ Tip
|
||||
<p>Check our other guides for details on these topics:</p>
|
||||
<ul>
|
||||
<li><a href="../docs/config.html">Configuration Guide</a> - Full configuration options</li>
|
||||
<li><a href="dataset-loading.qmd">Dataset Loading</a> - Loading datasets from various sources</li>
|
||||
<li><a href="../docs/dataset_loading.html">Dataset Loading</a> - Loading datasets from various sources</li>
|
||||
<li><a href="dataset-formats">Dataset Formats</a> - Working with different data formats</li>
|
||||
<li><a href="../docs/multi-gpu.html">Multi-GPU Training</a></li>
|
||||
<li><a href="../docs/multi-node.html">Multi-Node Training</a></li>
|
||||
|
||||
Reference in New Issue
Block a user