Buckets:

hf-doc-build
/

doc

Files

xet

hf-doc-build/doc / transformers /main /en /cpu.html

HuggingFaceDocBuilder

about 1 hour ago

download

raw

19 kB

<meta charset="utf-8" /><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;CPU&quot;,&quot;local&quot;:&quot;cpu&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Run inference&quot;,&quot;local&quot;:&quot;run-inference&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Train with Trainer&quot;,&quot;local&quot;:&quot;train-with-trainer&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Tune CPU performance&quot;,&quot;local&quot;:&quot;tune-cpu-performance&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Next steps&quot;,&quot;local&quot;:&quot;next-steps&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2}],&quot;depth&quot;:1}"/>
		<link href="/docs/transformers/main/en/_app/immutable/entry/start.Dg12X9ie.js" rel="modulepreload">
		<link href="/docs/transformers/main/en/_app/immutable/chunks/BKaGy-FC.js" rel="modulepreload">
		<link href="/docs/transformers/main/en/_app/immutable/chunks/DmjbnfDo.js" rel="modulepreload">
		<link href="/docs/transformers/main/en/_app/immutable/entry/app.DcrtE-99.js" rel="modulepreload">
		<link href="/docs/transformers/main/en/_app/immutable/chunks/HH-1qFgI.js" rel="modulepreload">
		<link href="/docs/transformers/main/en/_app/immutable/chunks/CcEgvo7V.js" rel="modulepreload">
		<link href="/docs/transformers/main/en/_app/immutable/chunks/DOLBYqcr.js" rel="modulepreload">
		<link href="/docs/transformers/main/en/_app/immutable/chunks/DsnmJJEf.js" rel="modulepreload">
		<link href="/docs/transformers/main/en/_app/immutable/nodes/0.Bm-4HD4u.js" rel="modulepreload">
		<link href="/docs/transformers/main/en/_app/immutable/chunks/DlCljDx1.js" rel="modulepreload">
		<link href="/docs/transformers/main/en/_app/immutable/nodes/7.CiX8So0Q.js" rel="modulepreload">
		<link href="/docs/transformers/main/en/_app/immutable/chunks/BCepyA0R.js" rel="modulepreload">
		<link href="/docs/transformers/main/en/_app/immutable/chunks/BWyUwo-H.js" rel="modulepreload">
		<!--cj616l--><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;CPU&quot;,&quot;local&quot;:&quot;cpu&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Run inference&quot;,&quot;local&quot;:&quot;run-inference&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Train with Trainer&quot;,&quot;local&quot;:&quot;train-with-trainer&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Tune CPU performance&quot;,&quot;local&quot;:&quot;tune-cpu-performance&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Next steps&quot;,&quot;local&quot;:&quot;next-steps&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2}],&quot;depth&quot;:1}"/><!---->
		<link href="/docs/transformers/main/en/_app/immutable/assets/0.tn0RQdqM.css" rel="modulepreload"> <!--[--><!--[0--><!--[--><!--[0--><!--[--><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> <!--[-1--><!--]--></div><!----> <!--[0--><h1 class="relative group"><a id="cpu" 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="#cpu"><span><svg 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>CPU</span></h1><!--]--><!----> <p>CPUs run Transformers models with the device string <code>cpu</code>, without an accelerator. They work for inference and training when GPUs aren’t available, or when you want a cost-effective setup.</p> <!--[1--><h2 class="relative group"><a id="run-inference" 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="#run-inference"><span><svg 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>Run inference</span></h2><!--]--><!----> <p>Models load on CPU by default when you don’t pass <code>device_map</code>, and <a href="/docs/transformers/main/en/main_classes/pipelines#transformers.pipeline">pipeline()</a> uses the CPU when no accelerator is available. Pass <code>device_map="cpu"</code> or <code>device="cpu"</code> to use the CPU on a machine that also has an accelerator.</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 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="language-py "><!----><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> pipeline

generator = pipeline(<span class="hljs-string">&quot;text-generation&quot;</span>, model=<span class="hljs-string">&quot;Qwen/Qwen3-0.6B&quot;</span>, device=<span class="hljs-string">&quot;cpu&quot;</span>)
<span class="hljs-built_in">print</span>(generator(<span class="hljs-string">&quot;CPUs are&quot;</span>, max_new_tokens=<span class="hljs-number">50</span>, do_sample=<span class="hljs-literal">False</span>)[<span class="hljs-number">0</span>][<span class="hljs-string">&quot;generated_text&quot;</span>])<!----></pre></div><!----> <!--[1--><h2 class="relative group"><a id="train-with-trainer" 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="#train-with-trainer"><span><svg 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>Train with Trainer</span></h2><!--]--><!----> <p>CPUs with bf16 support, such as recent Intel and AMD processors, can use bf16 mixed precision training with <a href="https://docs.pytorch.org/docs/stable/amp" rel="nofollow">PyTorch’s AMP</a>, which reduces memory usage and speeds up training.</p> <p>The examples below cover three setups, from a single CPU to several machines when one CPU is too slow:</p> <ul><li>a single CPU</li> <li>multiple processes on one machine (one per CPU socket)</li> <li>multiple processes across several machines</li></ul> <p>The single-node and multi-node examples launch processes with <code>mpirun</code> from <a href="https://www.intel.com/content/www/us/en/developer/tools/oneapi/mpi-library.html" rel="nofollow">Intel MPI</a>, part of the <a href="https://www.intel.com/content/www/us/en/developer/tools/oneapi/hpc-toolkit.html" rel="nofollow">Intel oneAPI HPC Toolkit</a>, and <a href="/docs/transformers/main/en/main_classes/trainer#transformers.Trainer">Trainer</a> runs DDP over the <code>gloo</code> backend.</p> <div class="flex space-x-2 items-center my-1.5 mr-8 h-7 !pl-0 -mx-3 md:mx-0"><!--[--><div class="flex items-center border rounded-lg px-1.5 py-1 leading-none select-none text-smd border-gray-800 bg-black dark:bg-gray-700 text-white">single CPU</div><div class="flex items-center border rounded-lg px-1.5 py-1 leading-none select-none text-smd text-gray-500 cursor-pointer opacity-90 hover:text-gray-700 dark:hover:text-gray-200 hover:shadow-sm">distributed CPU (single node)</div><div class="flex items-center border rounded-lg px-1.5 py-1 leading-none select-none text-smd text-gray-500 cursor-pointer opacity-90 hover:text-gray-700 dark:hover:text-gray-200 hover:shadow-sm">distributed CPU (multiple nodes)</div><!--]--></div> <div class="language-select"><!--[0--><p>Set <code>bf16=True</code> and <code>use_cpu=True</code> in <a href="/docs/transformers/main/en/main_classes/trainer#transformers.TrainingArguments">TrainingArguments</a>, or pass <code>--bf16</code> and <code>--use_cpu</code> to a training script. Prefer bf16 over fp16 on CPU because it’s more numerically stable.</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 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="language-py "><!----><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> TrainingArguments

training_args = TrainingArguments(
    output_dir=<span class="hljs-string">&quot;./outputs&quot;</span>,
    bf16=<span class="hljs-literal">True</span>,
    use_cpu=<span class="hljs-literal">True</span>,
)<!----></pre></div><!----><!----><!--]--><!----> <!--[-1--><!--]--><!----> <!--[-1--><!--]--><!----><!----></div><!----> <!--[1--><h2 class="relative group"><a id="tune-cpu-performance" 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="#tune-cpu-performance"><span><svg 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>Tune CPU performance</span></h2><!--]--><!----> <p>Preload a faster memory allocator and Intel’s OpenMP runtime to reduce overhead in CPU training. Install <code>google-perftools</code> (for <code>libtcmalloc</code>) and the <code>intel-openmp</code> Python package (for <code>libiomp5</code>), then add both libraries to the <code>LD_PRELOAD</code> environment variable before you launch training.</p> <p>Set <code>OMP_NUM_THREADS</code> to the number of cores each process can use. In a container or a cluster job with CPU limits, that’s the number of allocated CPU units, so leave some cores unallocated for system processes.</p> <!--[1--><h2 class="relative group"><a id="next-steps" 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="#next-steps"><span><svg 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>Next steps</span></h2><!--]--><!----> <ul><li>Read the <a href="https://huggingface.co/blog/intel-sapphire-rapids" rel="nofollow">Accelerating PyTorch Transformers with Intel Sapphire Rapids</a> blog post for a deeper look at BF16 performance on modern Intel hardware.</li> <li>Read <a href="./ddp">DDP</a> for more on the distributed strategy the multi-process examples use.</li> <li>Read the PyTorch <a href="https://docs.pytorch.org/docs/stable/notes/cpu_threading_torchscript_inference.html" rel="nofollow">CPU threading</a> notes to learn how <code>OMP_NUM_THREADS</code> affects performance.</li> <li>Read the Kubeflow <a href="https://www.kubeflow.org/docs/components/training/user-guides/pytorch/" rel="nofollow">PyTorchJob guide</a> to run distributed CPU training on Kubernetes.</li></ul> <a class="!text-gray-400 !no-underline text-sm flex items-center not-prose mt-4" href="https://github.com/huggingface/transformers/blob/main/docs/source/en/cpu.md" 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><span class="underline">Update</span> on GitHub</span></a><!----> <p></p><!--]--><!----><!--]--><!--]--><!--]--> <!--[-1--><!--]--><!--]-->
			
			<script>
				{
					__sveltekit_17blcf0 = {
						base: "/docs/transformers/main/en",
						assets: "/docs/transformers/main/en"
					};
					const element = document.currentScript.parentElement;
					Promise.all([
						import("/docs/transformers/main/en/_app/immutable/entry/start.Dg12X9ie.js"),
						import("/docs/transformers/main/en/_app/immutable/entry/app.DcrtE-99.js")
					]).then(([kit, app]) => {
						kit.start(app, element, {
							node_ids: [0, 7],
							data: [null,null],
							form: null,
							error: null
						});
					});
				}
			</script>
		

Xet Storage Details

Size:: 19 kB
Xet hash:: 2c90e825f3bb940eefa68d66c11911335aed70c954b25a433e0941d3d2a95b99

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

	<meta charset="utf-8" /><meta name="hf:doc:metadata" content="{"title":"CPU","local":"cpu","sections":[{"title":"Run inference","local":"run-inference","sections":[],"depth":2},{"title":"Train with Trainer","local":"train-with-trainer","sections":[],"depth":2},{"title":"Tune CPU performance","local":"tune-cpu-performance","sections":[],"depth":2},{"title":"Next steps","local":"next-steps","sections":[],"depth":2}],"depth":1}"/>
	<link href="/docs/transformers/main/en/_app/immutable/entry/start.Dg12X9ie.js" rel="modulepreload">
	<link href="/docs/transformers/main/en/_app/immutable/chunks/BKaGy-FC.js" rel="modulepreload">
	<link href="/docs/transformers/main/en/_app/immutable/chunks/DmjbnfDo.js" rel="modulepreload">
	<link href="/docs/transformers/main/en/_app/immutable/entry/app.DcrtE-99.js" rel="modulepreload">
	<link href="/docs/transformers/main/en/_app/immutable/chunks/HH-1qFgI.js" rel="modulepreload">
	<link href="/docs/transformers/main/en/_app/immutable/chunks/CcEgvo7V.js" rel="modulepreload">
	<link href="/docs/transformers/main/en/_app/immutable/chunks/DOLBYqcr.js" rel="modulepreload">
	<link href="/docs/transformers/main/en/_app/immutable/chunks/DsnmJJEf.js" rel="modulepreload">
	<link href="/docs/transformers/main/en/_app/immutable/nodes/0.Bm-4HD4u.js" rel="modulepreload">
	<link href="/docs/transformers/main/en/_app/immutable/chunks/DlCljDx1.js" rel="modulepreload">
	<link href="/docs/transformers/main/en/_app/immutable/nodes/7.CiX8So0Q.js" rel="modulepreload">
	<link href="/docs/transformers/main/en/_app/immutable/chunks/BCepyA0R.js" rel="modulepreload">
	<link href="/docs/transformers/main/en/_app/immutable/chunks/BWyUwo-H.js" rel="modulepreload">
	<!--cj616l--><meta name="hf:doc:metadata" content="{"title":"CPU","local":"cpu","sections":[{"title":"Run inference","local":"run-inference","sections":[],"depth":2},{"title":"Train with Trainer","local":"train-with-trainer","sections":[],"depth":2},{"title":"Tune CPU performance","local":"tune-cpu-performance","sections":[],"depth":2},{"title":"Next steps","local":"next-steps","sections":[],"depth":2}],"depth":1}"/><!---->
	<link href="/docs/transformers/main/en/_app/immutable/assets/0.tn0RQdqM.css" rel="modulepreload"> <!--[--><!--[0--><!--[--><!--[0--><!--[--><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> <!--[-1--><!--]--></div><!----> <!--[0--><h1 class="relative group"><a id="cpu" 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="#cpu"><span><svg 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>CPU</span></h1><!--]--><!----> <p>CPUs run Transformers models with the device string <code>cpu</code>, without an accelerator. They work for inference and training when GPUs aren’t available, or when you want a cost-effective setup.</p> <!--[1--><h2 class="relative group"><a id="run-inference" 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="#run-inference"><span><svg 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>Run inference</span></h2><!--]--><!----> <p>Models load on CPU by default when you don’t pass <code>device_map</code>, and <a href="/docs/transformers/main/en/main_classes/pipelines#transformers.pipeline">pipeline()</a> uses the CPU when no accelerator is available. Pass <code>device_map="cpu"</code> or <code>device="cpu"</code> to use the CPU on a machine that also has an accelerator.</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 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="language-py "><!----><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> pipeline

	generator = pipeline(<span class="hljs-string">"text-generation"</span>, model=<span class="hljs-string">"Qwen/Qwen3-0.6B"</span>, device=<span class="hljs-string">"cpu"</span>)
	<span class="hljs-built_in">print</span>(generator(<span class="hljs-string">"CPUs are"</span>, max_new_tokens=<span class="hljs-number">50</span>, do_sample=<span class="hljs-literal">False</span>)[<span class="hljs-number">0</span>][<span class="hljs-string">"generated_text"</span>])<!----></pre></div><!----> <!--[1--><h2 class="relative group"><a id="train-with-trainer" 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="#train-with-trainer"><span><svg 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>Train with Trainer</span></h2><!--]--><!----> <p>CPUs with bf16 support, such as recent Intel and AMD processors, can use bf16 mixed precision training with <a href="https://docs.pytorch.org/docs/stable/amp" rel="nofollow">PyTorch’s AMP</a>, which reduces memory usage and speeds up training.</p> <p>The examples below cover three setups, from a single CPU to several machines when one CPU is too slow:</p> <ul><li>a single CPU</li> <li>multiple processes on one machine (one per CPU socket)</li> <li>multiple processes across several machines</li></ul> <p>The single-node and multi-node examples launch processes with <code>mpirun</code> from <a href="https://www.intel.com/content/www/us/en/developer/tools/oneapi/mpi-library.html" rel="nofollow">Intel MPI</a>, part of the <a href="https://www.intel.com/content/www/us/en/developer/tools/oneapi/hpc-toolkit.html" rel="nofollow">Intel oneAPI HPC Toolkit</a>, and <a href="/docs/transformers/main/en/main_classes/trainer#transformers.Trainer">Trainer</a> runs DDP over the <code>gloo</code> backend.</p> <div class="flex space-x-2 items-center my-1.5 mr-8 h-7 !pl-0 -mx-3 md:mx-0"><!--[--><div class="flex items-center border rounded-lg px-1.5 py-1 leading-none select-none text-smd border-gray-800 bg-black dark:bg-gray-700 text-white">single CPU</div><div class="flex items-center border rounded-lg px-1.5 py-1 leading-none select-none text-smd text-gray-500 cursor-pointer opacity-90 hover:text-gray-700 dark:hover:text-gray-200 hover:shadow-sm">distributed CPU (single node)</div><div class="flex items-center border rounded-lg px-1.5 py-1 leading-none select-none text-smd text-gray-500 cursor-pointer opacity-90 hover:text-gray-700 dark:hover:text-gray-200 hover:shadow-sm">distributed CPU (multiple nodes)</div><!--]--></div> <div class="language-select"><!--[0--><p>Set <code>bf16=True</code> and <code>use_cpu=True</code> in <a href="/docs/transformers/main/en/main_classes/trainer#transformers.TrainingArguments">TrainingArguments</a>, or pass <code>--bf16</code> and <code>--use_cpu</code> to a training script. Prefer bf16 over fp16 on CPU because it’s more numerically stable.</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 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="language-py "><!----><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> TrainingArguments

	training_args = TrainingArguments(
	output_dir=<span class="hljs-string">"./outputs"</span>,
	bf16=<span class="hljs-literal">True</span>,
	use_cpu=<span class="hljs-literal">True</span>,
	)<!----></pre></div><!----><!----><!--]--><!----> <!--[-1--><!--]--><!----> <!--[-1--><!--]--><!----><!----></div><!----> <!--[1--><h2 class="relative group"><a id="tune-cpu-performance" 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="#tune-cpu-performance"><span><svg 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>Tune CPU performance</span></h2><!--]--><!----> <p>Preload a faster memory allocator and Intel’s OpenMP runtime to reduce overhead in CPU training. Install <code>google-perftools</code> (for <code>libtcmalloc</code>) and the <code>intel-openmp</code> Python package (for <code>libiomp5</code>), then add both libraries to the <code>LD_PRELOAD</code> environment variable before you launch training.</p> <p>Set <code>OMP_NUM_THREADS</code> to the number of cores each process can use. In a container or a cluster job with CPU limits, that’s the number of allocated CPU units, so leave some cores unallocated for system processes.</p> <!--[1--><h2 class="relative group"><a id="next-steps" 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="#next-steps"><span><svg 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>Next steps</span></h2><!--]--><!----> <ul><li>Read the <a href="https://huggingface.co/blog/intel-sapphire-rapids" rel="nofollow">Accelerating PyTorch Transformers with Intel Sapphire Rapids</a> blog post for a deeper look at BF16 performance on modern Intel hardware.</li> <li>Read <a href="./ddp">DDP</a> for more on the distributed strategy the multi-process examples use.</li> <li>Read the PyTorch <a href="https://docs.pytorch.org/docs/stable/notes/cpu_threading_torchscript_inference.html" rel="nofollow">CPU threading</a> notes to learn how <code>OMP_NUM_THREADS</code> affects performance.</li> <li>Read the Kubeflow <a href="https://www.kubeflow.org/docs/components/training/user-guides/pytorch/" rel="nofollow">PyTorchJob guide</a> to run distributed CPU training on Kubernetes.</li></ul> <a class="!text-gray-400 !no-underline text-sm flex items-center not-prose mt-4" href="https://github.com/huggingface/transformers/blob/main/docs/source/en/cpu.md" 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><span class="underline">Update</span> on GitHub</span></a><!----> <p></p><!--]--><!----><!--]--><!--]--><!--]--> <!--[-1--><!--]--><!--]-->

	<script>
	{
	__sveltekit_17blcf0 = {
	base: "/docs/transformers/main/en",
	assets: "/docs/transformers/main/en"
	};

	const element = document.currentScript.parentElement;

	Promise.all([
	import("/docs/transformers/main/en/_app/immutable/entry/start.Dg12X9ie.js"),
	import("/docs/transformers/main/en/_app/immutable/entry/app.DcrtE-99.js")
	]).then(([kit, app]) => {
	kit.start(app, element, {
	node_ids: [0, 7],
	data: [null,null],
	form: null,
	error: null
	});
	});
	}
	</script>