tocmo0nlord/axolotl
1
0
Fork 0
Code Issues Pull Requests Actions 3 Packages Projects Releases Wiki Activity

Built site for gh-pages

Browse Source
This commit is contained in:
Quarto GHA Workflow Runner
2025-09-16 18:58:53 +00:00
parent db626de56e
commit 421eea620c
209 changed files with 4822 additions and 3261 deletions
Download Patch File Download Diff File Expand all files Collapse all files

45
docs/api/utils.ctx_managers.sequence_parallel.html
View File

@@ -2,7 +2,7 @@
<html xmlns="http://www.w3.org/1999/xhtml" lang="en" xml:lang="en"><head>
<meta charset="utf-8">
<meta name="generator" content="quarto-1.7.34">
<meta name="generator" content="quarto-1.8.24">
<meta name="viewport" content="width=device-width, initial-scale=1.0, user-scalable=yes">
@@ -67,14 +67,15 @@ pre > code.sourceCode > span > a:first-child::before { text-decoration: underlin
<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/axe/axe-check.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-befe23ebd2f54d8af2c8a89d1a1611f1.css" rel="stylesheet" id="quarto-text-highlighting-styles">
<link href="../../site_libs/quarto-html/quarto-syntax-highlighting-dark-b651517ce65839d647a86e2780455cfb.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-e9895ec3143e9833a687747e8d39d226.min.css" rel="stylesheet" append-hash="true" id="quarto-bootstrap" data-mode="dark">
<link href="../../site_libs/bootstrap/bootstrap-f9d679a32da2b248d4ca48a0e58e089e.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,
@@ -124,7 +125,8 @@ gtag('config', 'G-9KYCVJBNMQ', { 'anonymize_ip': true});
<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">
<img src="../../image/axolotl_logo_digital_white.svg" alt="" class="navbar-logo light-content">
<img src="../../image/axolotl_logo_digital_white.svg" alt="" class="navbar-logo dark-content">
</a>
</div>
<div class="quarto-navbar-tools tools-wide tools-end">
@@ -150,6 +152,10 @@ gtag('config', 'G-9KYCVJBNMQ', { 'anonymize_ip': true});
<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="pt-lg-2 mt-2 text-left sidebar-header">
<a href="../../index.html" class="sidebar-logo-link">
</a>
</div>
<div class="sidebar-menu-container">
<ul class="list-unstyled mt-1">
<li class="sidebar-item">
@@ -544,7 +550,7 @@ gtag('config', 'G-9KYCVJBNMQ', { 'anonymize_ip': true});
</table>
<section id="axolotl.utils.ctx_managers.sequence_parallel.AllGatherWithGrad" class="level3">
<h3 class="anchored" data-anchor-id="axolotl.utils.ctx_managers.sequence_parallel.AllGatherWithGrad">AllGatherWithGrad</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.ctx_managers.sequence_parallel.AllGatherWithGrad()</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
<div class="code-copy-outer-scaffold"><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.ctx_managers.sequence_parallel.AllGatherWithGrad()</span></code></pre></div><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></div>
<p>Custom autograd function for all-gather to preserve gradients.</p>
<section id="methods" class="level4">
<h4 class="anchored" data-anchor-id="methods">Methods</h4>
@@ -568,10 +574,10 @@ gtag('config', 'G-9KYCVJBNMQ', { 'anonymize_ip': true});
</table>
<section id="axolotl.utils.ctx_managers.sequence_parallel.AllGatherWithGrad.backward" class="level5">
<h5 class="anchored" data-anchor-id="axolotl.utils.ctx_managers.sequence_parallel.AllGatherWithGrad.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.ctx_managers.sequence_parallel.AllGatherWithGrad.backward(</span>
<div class="code-copy-outer-scaffold"><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.ctx_managers.sequence_parallel.AllGatherWithGrad.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> grad_output,</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>
<span id="cb2-4"><a href="#cb2-4" aria-hidden="true" tabindex="-1"></a>)</span></code></pre></div><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></div>
<p>Backward pass for all-gather operation.</p>
<p>Extracts the gradient slice corresponding to this rank’s original input
from the full gradient tensor.</p>
@@ -635,11 +641,11 @@ from the full gradient tensor.</p>
</section>
<section id="axolotl.utils.ctx_managers.sequence_parallel.AllGatherWithGrad.forward" class="level5">
<h5 class="anchored" data-anchor-id="axolotl.utils.ctx_managers.sequence_parallel.AllGatherWithGrad.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.ctx_managers.sequence_parallel.AllGatherWithGrad.forward(</span>
<div class="code-copy-outer-scaffold"><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.ctx_managers.sequence_parallel.AllGatherWithGrad.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> input_tensor,</span>
<span id="cb3-4"><a href="#cb3-4" aria-hidden="true" tabindex="-1"></a> group,</span>
<span id="cb3-5"><a href="#cb3-5" aria-hidden="true" tabindex="-1"></a>)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
<span id="cb3-5"><a href="#cb3-5" aria-hidden="true" tabindex="-1"></a>)</span></code></pre></div><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></div>
<p>Forward pass of all-gather of data with sequence dimension.</p>
<section id="parameters-1" class="level6 doc-section doc-section-parameters">
<h6 class="doc-section doc-section-parameters anchored" data-anchor-id="parameters-1">Parameters</h6>
@@ -709,7 +715,7 @@ from the full gradient tensor.</p>
</section>
<section id="axolotl.utils.ctx_managers.sequence_parallel.SequenceParallelContextManager" class="level3">
<h3 class="anchored" data-anchor-id="axolotl.utils.ctx_managers.sequence_parallel.SequenceParallelContextManager">SequenceParallelContextManager</h3>
<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.ctx_managers.sequence_parallel.SequenceParallelContextManager(</span>
<div class="code-copy-outer-scaffold"><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.ctx_managers.sequence_parallel.SequenceParallelContextManager(</span>
<span id="cb4-2"><a href="#cb4-2" aria-hidden="true" tabindex="-1"></a> models,</span>
<span id="cb4-3"><a href="#cb4-3" aria-hidden="true" tabindex="-1"></a> context_parallel_size,</span>
<span id="cb4-4"><a href="#cb4-4" aria-hidden="true" tabindex="-1"></a> gradient_accumulation_steps,</span>
@@ -717,7 +723,7 @@ from the full gradient tensor.</p>
<span id="cb4-6"><a href="#cb4-6" aria-hidden="true" tabindex="-1"></a> heads_k_stride,</span>
<span id="cb4-7"><a href="#cb4-7" aria-hidden="true" tabindex="-1"></a> gather_outputs,</span>
<span id="cb4-8"><a href="#cb4-8" aria-hidden="true" tabindex="-1"></a> device_mesh<span class="op">=</span><span class="va">None</span>,</span>
<span id="cb4-9"><a href="#cb4-9" aria-hidden="true" tabindex="-1"></a>)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
<span id="cb4-9"><a href="#cb4-9" aria-hidden="true" tabindex="-1"></a>)</span></code></pre></div><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></div>
<p>Context manager for sequence parallelism operations.</p>
<p>This class provides a context that will automatically apply sequence parallelism
during model forward passes using a pre-forward hook, and gather outputs from
@@ -799,13 +805,13 @@ across the sequence parallelism group using a post-forward hook.</p>
</table>
<section id="axolotl.utils.ctx_managers.sequence_parallel.apply_sequence_parallelism" class="level3">
<h3 class="anchored" data-anchor-id="axolotl.utils.ctx_managers.sequence_parallel.apply_sequence_parallelism">apply_sequence_parallelism</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.ctx_managers.sequence_parallel.apply_sequence_parallelism(</span>
<div class="code-copy-outer-scaffold"><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.ctx_managers.sequence_parallel.apply_sequence_parallelism(</span>
<span id="cb5-2"><a href="#cb5-2" aria-hidden="true" tabindex="-1"></a> batch,</span>
<span id="cb5-3"><a href="#cb5-3" aria-hidden="true" tabindex="-1"></a> local_rank,</span>
<span id="cb5-4"><a href="#cb5-4" aria-hidden="true" tabindex="-1"></a> local_world_size,</span>
<span id="cb5-5"><a href="#cb5-5" aria-hidden="true" tabindex="-1"></a> gradient_accumulation_steps,</span>
<span id="cb5-6"><a href="#cb5-6" aria-hidden="true" tabindex="-1"></a> ring_attn_func,</span>
<span id="cb5-7"><a href="#cb5-7" aria-hidden="true" tabindex="-1"></a>)</span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
<span id="cb5-7"><a href="#cb5-7" aria-hidden="true" tabindex="-1"></a>)</span></code></pre></div><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></div>
<p>Apply sequence parallelism slicing to a batch.</p>
<p>Special handling is implemented for integer logits_to_keep, which indicates
to only keep the last N tokens in the sequence during generation.</p>
@@ -942,13 +948,14 @@ to only keep the last N tokens in the sequence during generation.</p>
e.clearSelection();
}
const getTextToCopy = function(trigger) {
const codeEl = trigger.previousElementSibling.cloneNode(true);
for (const childEl of codeEl.children) {
if (isCodeAnnotation(childEl)) {
childEl.remove();
}
const outerScaffold = trigger.parentElement.cloneNode(true);
const codeEl = outerScaffold.querySelector('code');
for (const childEl of codeEl.children) {
if (isCodeAnnotation(childEl)) {
childEl.remove();
}
return codeEl.innerText;
}
return codeEl.innerText;
}
const clipboard = new window.ClipboardJS('.code-copy-button:not([data-in-quarto-modal])', {
text: getTextToCopy