Hugging Face's logo

Buckets:
hf-doc-build
/
doc-dev

Files
xet
hf-doc-build/doc-dev / lerobot /pr_3313 /en /multi_gpu_training.html
HuggingFaceDocBuilder's picture
HuggingFaceDocBuilder
download
raw
32.8 kB
<meta charset="utf-8" /><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;Multi-GPU Training&quot;,&quot;local&quot;:&quot;multi-gpu-training&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Installation&quot;,&quot;local&quot;:&quot;installation&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Training with Multiple GPUs&quot;,&quot;local&quot;:&quot;training-with-multiple-gpus&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Option 1: Without config (specify parameters directly)&quot;,&quot;local&quot;:&quot;option-1-without-config-specify-parameters-directly&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3},{&quot;title&quot;:&quot;Option 2: Using accelerate config&quot;,&quot;local&quot;:&quot;option-2-using-accelerate-config&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3}],&quot;depth&quot;:2},{&quot;title&quot;:&quot;How It Works&quot;,&quot;local&quot;:&quot;how-it-works&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Learning Rate and Training Steps Scaling&quot;,&quot;local&quot;:&quot;learning-rate-and-training-steps-scaling&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Why No Automatic Scaling?&quot;,&quot;local&quot;:&quot;why-no-automatic-scaling&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3},{&quot;title&quot;:&quot;When and How to Scale&quot;,&quot;local&quot;:&quot;when-and-how-to-scale&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3}],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Notes&quot;,&quot;local&quot;:&quot;notes&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2}],&quot;depth&quot;:1}">
 <link href="/docs/lerobot/pr_3313/en/_app/immutable/assets/0.e3b0c442.css" rel="modulepreload">
 <link rel="modulepreload" href="/docs/lerobot/pr_3313/en/_app/immutable/entry/start.d3f1c0f3.js">
 <link rel="modulepreload" href="/docs/lerobot/pr_3313/en/_app/immutable/chunks/scheduler.eb244325.js">
 <link rel="modulepreload" href="/docs/lerobot/pr_3313/en/_app/immutable/chunks/singletons.1f33814c.js">
 <link rel="modulepreload" href="/docs/lerobot/pr_3313/en/_app/immutable/chunks/index.3c23fb4b.js">
 <link rel="modulepreload" href="/docs/lerobot/pr_3313/en/_app/immutable/chunks/paths.17f05d75.js">
 <link rel="modulepreload" href="/docs/lerobot/pr_3313/en/_app/immutable/entry/app.04bb7687.js">
 <link rel="modulepreload" href="/docs/lerobot/pr_3313/en/_app/immutable/chunks/preload-helper.b00aacbc.js">
 <link rel="modulepreload" href="/docs/lerobot/pr_3313/en/_app/immutable/chunks/index.3fe63ad3.js">
 <link rel="modulepreload" href="/docs/lerobot/pr_3313/en/_app/immutable/nodes/0.07fbe93e.js">
 <link rel="modulepreload" href="/docs/lerobot/pr_3313/en/_app/immutable/chunks/each.e59479a4.js">
 <link rel="modulepreload" href="/docs/lerobot/pr_3313/en/_app/immutable/nodes/36.f15cb129.js">
 <link rel="modulepreload" href="/docs/lerobot/pr_3313/en/_app/immutable/chunks/CopyLLMTxtMenu.d0c64540.js">
 <link rel="modulepreload" href="/docs/lerobot/pr_3313/en/_app/immutable/chunks/MermaidChart.svelte_svelte_type_style_lang.6453902c.js">
 <link rel="modulepreload" href="/docs/lerobot/pr_3313/en/_app/immutable/chunks/CodeBlock.48dd2cc2.js"><!-- HEAD_svelte-u9bgzb_START --><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;Multi-GPU Training&quot;,&quot;local&quot;:&quot;multi-gpu-training&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Installation&quot;,&quot;local&quot;:&quot;installation&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Training with Multiple GPUs&quot;,&quot;local&quot;:&quot;training-with-multiple-gpus&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Option 1: Without config (specify parameters directly)&quot;,&quot;local&quot;:&quot;option-1-without-config-specify-parameters-directly&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3},{&quot;title&quot;:&quot;Option 2: Using accelerate config&quot;,&quot;local&quot;:&quot;option-2-using-accelerate-config&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3}],&quot;depth&quot;:2},{&quot;title&quot;:&quot;How It Works&quot;,&quot;local&quot;:&quot;how-it-works&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Learning Rate and Training Steps Scaling&quot;,&quot;local&quot;:&quot;learning-rate-and-training-steps-scaling&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Why No Automatic Scaling?&quot;,&quot;local&quot;:&quot;why-no-automatic-scaling&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3},{&quot;title&quot;:&quot;When and How to Scale&quot;,&quot;local&quot;:&quot;when-and-how-to-scale&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3}],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Notes&quot;,&quot;local&quot;:&quot;notes&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2}],&quot;depth&quot;:1}"><!-- HEAD_svelte-u9bgzb_END --> <p></p> <div class="items-center shrink-0 min-w-[100px] max-sm:min-w-[50px] justify-end ml-auto flex" style="float: right; margin-left: 10px; display: inline-flex; position: relative; z-index: 10;"><div class="inline-flex rounded-md max-sm:rounded-sm"><button class="inline-flex items-center gap-1 h-7 max-sm:h-7 px-2 max-sm:px-1.5 text-sm font-medium text-gray-800 border border-r-0 rounded-l-md max-sm:rounded-l-sm border-gray-200 bg-white hover:shadow-inner dark:border-gray-850 dark:bg-gray-950 dark:text-gray-200 dark:hover:bg-gray-800" aria-live="polite"><span class="inline-flex items-center justify-center rounded-md p-0.5 max-sm:p-0 hover:text-gray-800 dark:hover:text-gray-200"><svg class="sm:size-3.5 size-3" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg></span> <span>Copy page</span></button> <button class="inline-flex items-center justify-center w-6 max-sm:w-5 h-7 max-sm:h-7 disabled:pointer-events-none text-sm text-gray-500 hover:text-gray-700 dark:hover:text-white rounded-r-md max-sm:rounded-r-sm border border-l transition border-gray-200 bg-white hover:shadow-inner dark:border-gray-850 dark:bg-gray-950 dark:text-gray-200 dark:hover:bg-gray-800" aria-haspopup="menu" aria-expanded="false" aria-label="Open copy menu"><svg class="transition-transform text-gray-400 overflow-visible sm:size-3.5 size-3 rotate-0" width="1em" height="1em" viewBox="0 0 12 7" fill="none" xmlns="http://www.w3.org/2000/svg"><path d="M1 1L6 6L11 1" stroke="currentColor"></path></svg></button></div> </div> <h1 class="relative group"><a id="multi-gpu-training" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#multi-gpu-training"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>Multi-GPU Training</span></h1> <p data-svelte-h="svelte-1q21bfz">This guide shows you how to train policies on multiple GPUs using <a href="https://huggingface.co/docs/accelerate" rel="nofollow">Hugging Face Accelerate</a>.</p> <h2 class="relative group"><a id="installation" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#installation"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>Installation</span></h2> <p data-svelte-h="svelte-zyh6gt"><code>accelerate</code> is included in the <code>training</code> extra. Install it with:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->pip install <span class="hljs-string">&#x27;lerobot[training]&#x27;</span><!-- HTML_TAG_END --></pre></div> <h2 class="relative group"><a id="training-with-multiple-gpus" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#training-with-multiple-gpus"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>Training with Multiple GPUs</span></h2> <p data-svelte-h="svelte-1vye7mj">You can launch training in two ways:</p> <h3 class="relative group"><a id="option-1-without-config-specify-parameters-directly" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#option-1-without-config-specify-parameters-directly"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>Option 1: Without config (specify parameters directly)</span></h3> <p data-svelte-h="svelte-3gdn0b">You can specify all parameters directly in the command without running <code>accelerate config</code>:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->accelerate launch \
 --multi_gpu \
 --num_processes=2 \
 $(<span class="hljs-built_in">which</span> lerobot-train) \
 --dataset.repo_id=<span class="hljs-variable">${HF_USER}</span>/my_dataset \
 --policy.type=act \
 --policy.repo_id=<span class="hljs-variable">${HF_USER}</span>/my_trained_policy \
 --output_dir=outputs/train/act_multi_gpu \
 --job_name=act_multi_gpu \
 --wandb.enable=<span class="hljs-literal">true</span><!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-f2fa6v"><strong>Key accelerate parameters:</strong></p> <ul data-svelte-h="svelte-aqgwg9"><li><code>--multi_gpu</code>: Enable multi-GPU training</li> <li><code>--num_processes=2</code>: Number of GPUs to use</li> <li><code>--mixed_precision=fp16</code>: Use fp16 mixed precision (or <code>bf16</code> if supported)</li></ul> <h3 class="relative group"><a id="option-2-using-accelerate-config" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#option-2-using-accelerate-config"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>Option 2: Using accelerate config</span></h3> <p data-svelte-h="svelte-1b40wb9">If you prefer to save your configuration, you can optionally configure accelerate for your hardware setup by running:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->accelerate config<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1plecjh">This interactive setup will ask you questions about your training environment (number of GPUs, mixed precision settings, etc.) and saves the configuration for future use. For a simple multi-GPU setup on a single machine, you can use these recommended settings:</p> <ul data-svelte-h="svelte-1skdlw3"><li>Compute environment: This machine</li> <li>Number of machines: 1</li> <li>Number of processes: (number of GPUs you want to use)</li> <li>GPU ids to use: (leave empty to use all)</li> <li>Mixed precision: fp16 or bf16 (recommended for faster training)</li></ul> <p data-svelte-h="svelte-p23sye">Then launch training with:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->accelerate launch $(<span class="hljs-built_in">which</span> lerobot-train) \
 --dataset.repo_id=<span class="hljs-variable">${HF_USER}</span>/my_dataset \
 --policy.type=act \
 --policy.repo_id=<span class="hljs-variable">${HF_USER}</span>/my_trained_policy \
 --output_dir=outputs/train/act_multi_gpu \
 --job_name=act_multi_gpu \
 --wandb.enable=<span class="hljs-literal">true</span><!-- HTML_TAG_END --></pre></div> <h2 class="relative group"><a id="how-it-works" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#how-it-works"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>How It Works</span></h2> <p data-svelte-h="svelte-c2blf5">When you launch training with accelerate:</p> <ol data-svelte-h="svelte-1tbs832"><li><strong>Automatic detection</strong>: LeRobot automatically detects if it’s running under accelerate</li> <li><strong>Data distribution</strong>: Your batch is automatically split across GPUs</li> <li><strong>Gradient synchronization</strong>: Gradients are synchronized across GPUs during backpropagation</li> <li><strong>Single process logging</strong>: Only the main process logs to wandb and saves checkpoints</li></ol> <h2 class="relative group"><a id="learning-rate-and-training-steps-scaling" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#learning-rate-and-training-steps-scaling"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>Learning Rate and Training Steps Scaling</span></h2> <p data-svelte-h="svelte-rbj0qr"><strong>Important:</strong> LeRobot does <strong>NOT</strong> automatically scale learning rates or training steps based on the number of GPUs. This gives you full control over your training hyperparameters.</p> <h3 class="relative group"><a id="why-no-automatic-scaling" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#why-no-automatic-scaling"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>Why No Automatic Scaling?</span></h3> <p data-svelte-h="svelte-15zrztt">Many distributed training frameworks automatically scale the learning rate by the number of GPUs (e.g., <code>lr = base_lr × num_gpus</code>).
However, LeRobot keeps the learning rate exactly as you specify it.</p> <h3 class="relative group"><a id="when-and-how-to-scale" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#when-and-how-to-scale"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>When and How to Scale</span></h3> <p data-svelte-h="svelte-6v0wu4">If you want to scale your hyperparameters when using multiple GPUs, you should do it manually:</p> <p data-svelte-h="svelte-1ppb592"><strong>Learning Rate Scaling:</strong></p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START --><span class="hljs-comment"># Example: 2 GPUs with linear LR scaling</span>
<span class="hljs-comment"># Base LR: 1e-4, with 2 GPUs -&gt; 2e-4</span>
accelerate launch --num_processes=2 $(<span class="hljs-built_in">which</span> lerobot-train) \
 --optimizer.lr=2e-4 \
 --dataset.repo_id=lerobot/pusht \
 --policy=act<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-lmqw9n"><strong>Training Steps Scaling:</strong></p> <p data-svelte-h="svelte-z9u3ey">Since the effective batch size <code>bs</code> increases with multiple GPUs (batch_size × num_gpus), you may want to reduce the number of training steps proportionally:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START --><span class="hljs-comment"># Example: 2 GPUs with effective batch size 2x larger</span>
<span class="hljs-comment"># Original: batch_size=8, steps=100000</span>
<span class="hljs-comment"># With 2 GPUs: batch_size=8 (16 in total), steps=50000</span>
accelerate launch --num_processes=2 $(<span class="hljs-built_in">which</span> lerobot-train) \
 --batch_size=8 \
 --steps=50000 \
 --dataset.repo_id=lerobot/pusht \
 --policy=act<!-- HTML_TAG_END --></pre></div> <h2 class="relative group"><a id="notes" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#notes"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>Notes</span></h2> <ul data-svelte-h="svelte-1nirqzy"><li>The <code>--policy.use_amp</code> flag in <code>lerobot-train</code> is only used when <strong>not</strong> running with accelerate. When using accelerate, mixed precision is controlled by accelerate’s configuration.</li> <li>Training logs, checkpoints, and hub uploads are only done by the main process to avoid conflicts. Non-main processes have console logging disabled to prevent duplicate output.</li> <li>The effective batch size is <code>batch_size × num_gpus</code>. If you use 4 GPUs with <code>--batch_size=8</code>, your effective batch size is 32.</li> <li>Learning rate scheduling is handled correctly across multiple processes—LeRobot sets <code>step_scheduler_with_optimizer=False</code> to prevent accelerate from adjusting scheduler steps based on the number of processes.</li> <li>When saving or pushing models, LeRobot automatically unwraps the model from accelerate’s distributed wrapper to ensure compatibility.</li> <li>WandB integration automatically initializes only on the main process, preventing multiple runs from being created.</li></ul> <p data-svelte-h="svelte-qqmx1m">For more advanced configurations and troubleshooting, see the <a href="https://huggingface.co/docs/accelerate" rel="nofollow">Accelerate documentation</a>. If you want to learn more about how to train on a large number of GPUs, checkout this awesome guide: <a href="https://huggingface.co/spaces/nanotron/ultrascale-playbook" rel="nofollow">Ultrascale Playbook</a>.</p> <a class="!text-gray-400 !no-underline text-sm flex items-center not-prose mt-4" href="https://github.com/huggingface/lerobot/blob/main/docs/source/multi_gpu_training.mdx" target="_blank"><svg class="mr-1" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M31,16l-7,7l-1.41-1.41L28.17,16l-5.58-5.59L24,9l7,7z"></path><path d="M1,16l7-7l1.41,1.41L3.83,16l5.58,5.59L8,23l-7-7z"></path><path d="M12.419,25.484L17.639,6.552l1.932,0.518L14.351,26.002z"></path></svg> <span data-svelte-h="svelte-zjs2n5"><span class="underline">Update</span> on GitHub</span></a> <p></p>
<script>
 {
 __sveltekit_9kza6s = {
 assets: "/docs/lerobot/pr_3313/en",
 base: "/docs/lerobot/pr_3313/en",
 env: {}
 };
 const element = document.currentScript.parentElement;
 const data = [null,null];
 Promise.all([
 import("/docs/lerobot/pr_3313/en/_app/immutable/entry/start.d3f1c0f3.js"),
 import("/docs/lerobot/pr_3313/en/_app/immutable/entry/app.04bb7687.js")
 ]).then(([kit, app]) => {
 kit.start(app, element, {
 node_ids: [0, 36],
 data,
 form: null,
 error: null
 });
 });
 }
 </script>

Xet Storage Details

Size:
32.8 kB
·
Xet hash:
e329d05fb5f215fed7147d11c965c96535eb884f98142511f40826227feccde3

Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.