diffusers/main/ko/training/distributed_inference.html

<meta charset="utf-8" /><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;여러 GPU를 사용한 분산 추론&quot;,&quot;local&quot;:&quot;여러-gpu를-사용한-분산-추론&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;🤗 Accelerate&quot;,&quot;local&quot;:&quot;-accelerate&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Pytoerch 분산&quot;,&quot;local&quot;:&quot;pytoerch-분산&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2}],&quot;depth&quot;:1}">
		<link href="/docs/diffusers/main/ko/_app/immutable/assets/0.e3b0c442.css" rel="modulepreload">
		<link rel="modulepreload" href="/docs/diffusers/main/ko/_app/immutable/entry/start.df0b2e2f.js">
		<link rel="modulepreload" href="/docs/diffusers/main/ko/_app/immutable/chunks/scheduler.94020406.js">
		<link rel="modulepreload" href="/docs/diffusers/main/ko/_app/immutable/chunks/singletons.4482ca3c.js">
		<link rel="modulepreload" href="/docs/diffusers/main/ko/_app/immutable/chunks/index.8b553f6b.js">
		<link rel="modulepreload" href="/docs/diffusers/main/ko/_app/immutable/chunks/paths.1334326c.js">
		<link rel="modulepreload" href="/docs/diffusers/main/ko/_app/immutable/entry/app.4378a265.js">
		<link rel="modulepreload" href="/docs/diffusers/main/ko/_app/immutable/chunks/index.a08c8d92.js">
		<link rel="modulepreload" href="/docs/diffusers/main/ko/_app/immutable/nodes/0.96378992.js">
		<link rel="modulepreload" href="/docs/diffusers/main/ko/_app/immutable/chunks/each.e59479a4.js">
		<link rel="modulepreload" href="/docs/diffusers/main/ko/_app/immutable/nodes/25.d70cbdd3.js">
		<link rel="modulepreload" href="/docs/diffusers/main/ko/_app/immutable/chunks/Tip.3b0aeee8.js">
		<link rel="modulepreload" href="/docs/diffusers/main/ko/_app/immutable/chunks/CodeBlock.b23cf525.js">
		<link rel="modulepreload" href="/docs/diffusers/main/ko/_app/immutable/chunks/EditOnGithub.b1bceb47.js"><!-- HEAD_svelte-u9bgzb_START --><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;여러 GPU를 사용한 분산 추론&quot;,&quot;local&quot;:&quot;여러-gpu를-사용한-분산-추론&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;🤗 Accelerate&quot;,&quot;local&quot;:&quot;-accelerate&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Pytoerch 분산&quot;,&quot;local&quot;:&quot;pytoerch-분산&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2}],&quot;depth&quot;:1}"><!-- HEAD_svelte-u9bgzb_END -->      <p></p>  <h1 class="relative group"><a id="여러-gpu를-사용한-분산-추론" 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="#여러-gpu를-사용한-분산-추론"><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>여러 GPU를 사용한 분산 추론</span></h1> <p data-svelte-h="svelte-1pju33">분산 설정에서는 여러 개의 프롬프트를 동시에 생성할 때 유용한 🤗 <a href="https://huggingface.co/docs/accelerate/index" rel="nofollow">Accelerate</a> 또는 <a href="https://pytorch.org/tutorials/beginner/dist_overview.html" rel="nofollow">PyTorch Distributed</a>를 사용하여 여러 GPU에서 추론을 실행할 수 있습니다.</p> <p data-svelte-h="svelte-mr3tus">이 가이드에서는 분산 추론을 위해 🤗 Accelerate와 PyTorch Distributed를 사용하는 방법을 보여드립니다.</p>  <h2 class="relative group"><a id="-accelerate" 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="#-accelerate"><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>🤗 Accelerate</span></h2> <p data-svelte-h="svelte-gk0xcz">🤗 <a href="https://huggingface.co/docs/accelerate/index" rel="nofollow">Accelerate</a>는 분산 설정에서 추론을 쉽게 훈련하거나 실행할 수 있도록 설계된 라이브러리입니다. 분산 환경 설정 프로세스를 간소화하여 PyTorch 코드에 집중할 수 있도록 해줍니다.</p> <p data-svelte-h="svelte-t8zt9f">시작하려면 Python 파일을 생성하고 <code>accelerate.PartialState</code>를 초기화하여 분산 환경을 생성하면, 설정이 자동으로 감지되므로 <code>rank</code> 또는 <code>world_size</code>를 명시적으로 정의할 필요가 없습니다. [‘DiffusionPipeline`]을 <code>distributed_state.device</code>로 이동하여 각 프로세스에 GPU를 할당합니다.</p> <p data-svelte-h="svelte-1wde0xg">이제 컨텍스트 관리자로 <code>split_between_processes</code> 유틸리티를 사용하여 프로세스 수에 따라 프롬프트를 자동으로 분배합니다.</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-keyword">from</span> accelerate <span class="hljs-keyword">import</span> PartialState
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> DiffusionPipeline

pipeline = DiffusionPipeline.from_pretrained(<span class="hljs-string">&quot;runwayml/stable-diffusion-v1-5&quot;</span>, torch_dtype=torch.float16)
distributed_state = PartialState()
pipeline.to(distributed_state.device)

<span class="hljs-keyword">with</span> distributed_state.split_between_processes([<span class="hljs-string">&quot;a dog&quot;</span>, <span class="hljs-string">&quot;a cat&quot;</span>]) <span class="hljs-keyword">as</span> prompt:
    result = pipeline(prompt).images[<span class="hljs-number">0</span>]
    result.save(<span class="hljs-string">f&quot;result_<span class="hljs-subst">{distributed_state.process_index}</span>.png&quot;</span>)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1ohh8as">Use the <code>--num_processes</code> argument to specify the number of GPUs to use, and call <code>accelerate launch</code> to run the script:</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 run_distributed.py --num_processes=2<!-- HTML_TAG_END --></pre></div>  <div class="course-tip  bg-gradient-to-br dark:bg-gradient-to-r before:border-green-500 dark:before:border-green-800 from-green-50 dark:from-gray-900 to-white dark:to-gray-950 border border-green-50 text-green-700 dark:text-gray-400">자세한 내용은 [🤗 Accelerate를 사용한 분산 추론](https://huggingface.co/docs/accelerate/en/usage_guides/distributed_inference#distributed-inference-with-accelerate) 가이드를 참조하세요.</div>  <h2 class="relative group"><a id="pytoerch-분산" 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="#pytoerch-분산"><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>Pytoerch 분산</span></h2> <p data-svelte-h="svelte-db33v7">PyTorch는 데이터 병렬 처리를 가능하게 하는 <a href="https://pytorch.org/docs/stable/generated/torch.nn.parallel.DistributedDataParallel.html" rel="nofollow"><code>DistributedDataParallel</code></a>을 지원합니다.</p> <p data-svelte-h="svelte-td2orp">시작하려면 Python 파일을 생성하고 <code>torch.distributed</code> 및 <code>torch.multiprocessing</code>을 임포트하여 분산 프로세스 그룹을 설정하고 각 GPU에서 추론용 프로세스를 생성합니다. 그리고 <code>DiffusionPipeline</code>도 초기화해야 합니다:</p> <p data-svelte-h="svelte-2xw9l6">확산 파이프라인을 <code>rank</code>로 이동하고 <code>get_rank</code>를 사용하여 각 프로세스에 GPU를 할당하면 각 프로세스가 다른 프롬프트를 처리합니다:</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-keyword">import</span> torch
<span class="hljs-keyword">import</span> torch.distributed <span class="hljs-keyword">as</span> dist
<span class="hljs-keyword">import</span> torch.multiprocessing <span class="hljs-keyword">as</span> mp

<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> DiffusionPipeline

sd = DiffusionPipeline.from_pretrained(<span class="hljs-string">&quot;runwayml/stable-diffusion-v1-5&quot;</span>, torch_dtype=torch.float16)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-60cuu7">사용할 백엔드 유형, 현재 프로세스의 <code>rank</code>, <code>world_size</code> 또는 참여하는 프로세스 수로 분산 환경 생성을 처리하는 함수<code>init_process_group</code>를 만들어 추론을 실행해야 합니다.</p> <p data-svelte-h="svelte-1xq31ih">2개의 GPU에서 추론을 병렬로 실행하는 경우 <code>world_size</code>는 2입니다.</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-keyword">def</span> <span class="hljs-title function_">run_inference</span>(<span class="hljs-params">rank, world_size</span>):
    dist.init_process_group(<span class="hljs-string">&quot;nccl&quot;</span>, rank=rank, world_size=world_size)

    sd.to(rank)

    <span class="hljs-keyword">if</span> torch.distributed.get_rank() == <span class="hljs-number">0</span>:
        prompt = <span class="hljs-string">&quot;a dog&quot;</span>
    <span class="hljs-keyword">elif</span> torch.distributed.get_rank() == <span class="hljs-number">1</span>:
        prompt = <span class="hljs-string">&quot;a cat&quot;</span>

    image = sd(prompt).images[<span class="hljs-number">0</span>]
    image.save(<span class="hljs-string">f&quot;./<span class="hljs-subst">{<span class="hljs-string">&#x27;_&#x27;</span>.join(prompt)}</span>.png&quot;</span>)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1hoikpv">분산 추론을 실행하려면 <a href="https://pytorch.org/docs/stable/multiprocessing.html#torch.multiprocessing.spawn" rel="nofollow"><code>mp.spawn</code></a>을 호출하여 <code>world_size</code>에 정의된 GPU 수에 대해 <code>run_inference</code> 함수를 실행합니다:</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-keyword">def</span> <span class="hljs-title function_">main</span>():
    world_size = <span class="hljs-number">2</span>
    mp.spawn(run_inference, args=(world_size,), nprocs=world_size, join=<span class="hljs-literal">True</span>)


<span class="hljs-keyword">if</span> __name__ == <span class="hljs-string">&quot;__main__&quot;</span>:
    main()<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1g25qdf">추론 스크립트를 완료했으면 <code>--nproc_per_node</code> 인수를 사용하여 사용할 GPU 수를 지정하고 <code>torchrun</code>을 호출하여 스크립트를 실행합니다:</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 -->torchrun run_distributed.py --nproc_per_node=2<!-- HTML_TAG_END --></pre></div> <a class="!text-gray-400 !no-underline text-sm flex items-center not-prose mt-4" href="https://github.com/huggingface/diffusers/blob/main/docs/source/ko/training/distributed_inference.md" target="_blank"><span data-svelte-h="svelte-1kd6by1">&lt;</span> <span data-svelte-h="svelte-x0xyl0">&gt;</span> <span data-svelte-h="svelte-1dajgef"><span class="underline ml-1.5">Update</span> on GitHub</span></a>  <p></p> 
			
			<script>
				{
					__sveltekit_bmkzir = {
						assets: "/docs/diffusers/main/ko",
						base: "/docs/diffusers/main/ko",
						env: {}
					};

					const element = document.currentScript.parentElement;

					const data = [null,null];

					Promise.all([
						import("/docs/diffusers/main/ko/_app/immutable/entry/start.df0b2e2f.js"),
						import("/docs/diffusers/main/ko/_app/immutable/entry/app.4378a265.js")
					]).then(([kit, app]) => {
						kit.start(app, element, {
							node_ids: [0, 25],
							data,
							form: null,
							error: null
						});
					});
				}
			</script>