Buckets:

hf-doc-build
/

doc

Files
xet
hf-doc-build/doc / diffusers /main /ko /optimization /pruna.html
mishig
about 13 hours ago
download
raw
31.3 kB
<meta charset="utf-8" /><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;Pruna&quot;,&quot;local&quot;:&quot;pruna&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;설치&quot;,&quot;local&quot;:&quot;installation&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Diffusers 모델 최적화&quot;,&quot;local&quot;:&quot;optimize-diffusers-models&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Diffusers 모델 평가 및 벤치마킹&quot;,&quot;local&quot;:&quot;evaluate-and-benchmark-diffusers-models&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;참고&quot;,&quot;local&quot;:&quot;reference&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2}],&quot;depth&quot;:1}"/>
		<link href="/docs/diffusers/main/en/_app/immutable/entry/start.CM-2NerT.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/CTWHa_kr.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/BLi2et3c.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/entry/app.DrYcqW7m.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/BeiZzCMo.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/DnVIJfC5.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/DsnmJJEf.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/t03vT2vk.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/nodes/0.DdeITYyU.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/DAwShlVY.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/nodes/11.CufzJa4R.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/8Bi71p98.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/LdXCdsHQ.js" rel="modulepreload">
		<!--15jkk7f--><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;Pruna&quot;,&quot;local&quot;:&quot;pruna&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;Optimize Diffusers models&quot;,&quot;local&quot;:&quot;optimize-diffusers-models&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Evaluate and benchmark Diffusers models&quot;,&quot;local&quot;:&quot;evaluate-and-benchmark-diffusers-models&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Reference&quot;,&quot;local&quot;:&quot;reference&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2}],&quot;depth&quot;:1}"/><!---->
		<link href="/docs/diffusers/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: inline-end; margin-inline-start: 10px; display: inline-flex; position: relative; z-index: 10;"><div dir="ltr" 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 dir="auto">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="pruna" 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="#pruna"><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><!---->Pruna<!----></span></h1><!--]--><!----> <p><a href="https://github.com/PrunaAI/pruna" rel="nofollow">Pruna</a>는 추론을 가속화하고 메모리 사용량을 줄이기 위해 양자화, 프루닝, 캐싱, 컴파일 등 다양한 최적화 방법을 제공하는 모델 최적화 프레임워크입니다. 최적화 방법의 일반적인 개요는 아래에 나와 있습니다.</p> <table><thead><tr><th>기법</th><th>설명</th><th align="center">속도</th><th align="center">메모리</th><th align="center">품질</th></tr></thead><tbody><tr><td><code>batcher</code></td><td>여러 입력을 한 그룹으로 묶어 동시에 처리함으로써 계산 효율성을 높이고 처리 시간을 줄입니다.</td><td align="center">✅</td><td align="center">❌</td><td align="center">➖</td></tr><tr><td><code>cacher</code></td><td>계산의 중간 결과를 저장하여 후속 연산의 속도를 높입니다.</td><td align="center">✅</td><td align="center">➖</td><td align="center">➖</td></tr><tr><td><code>compiler</code></td><td>특정 하드웨어에 대한 지침으로 모델을 최적화합니다.</td><td align="center">✅</td><td align="center">➖</td><td align="center">➖</td></tr><tr><td><code>distiller</code></td><td>더 크고 복잡한 모델을 모방하도록 더 작고 단순한 모델을 학습합니다.</td><td align="center">✅</td><td align="center">✅</td><td align="center">❌</td></tr><tr><td><code>quantizer</code></td><td>가중치와 활성화의 정밀도를 낮추어 메모리 요구량을 줄입니다.</td><td align="center">✅</td><td align="center">✅</td><td align="center">❌</td></tr><tr><td><code>pruner</code></td><td>중요도가 낮거나 중복된 연결과 뉴런을 제거하여 더 희소하고 효율적인 네트워크를 생성합니다.</td><td align="center">✅</td><td align="center">✅</td><td align="center">❌</td></tr><tr><td><code>recoverer</code></td><td>압축 후 모델의 성능을 복원합니다.</td><td align="center">➖</td><td align="center">➖</td><td align="center">✅</td></tr><tr><td><code>factorizer</code></td><td>Factorization는 여러 작은 행렬 곱셈을 하나의 큰 융합된 연산으로 배치합니다.</td><td align="center">✅</td><td align="center">➖</td><td align="center">➖</td></tr><tr><td><code>enhancer</code></td><td>노이즈 제거 또는 업스케일링과 같은 후처리 알고리즘을 적용하여 모델 출력을 향상시킵니다.</td><td align="center">❌</td><td align="center">-</td><td align="center">✅</td></tr></tbody></table> <p>✅ (개선), ➖ (대략 동일), ❌ (악화)</p> <p><a href="https://docs.pruna.ai/en/stable/docs_pruna/user_manual/configure.html#configure-algorithms" rel="nofollow">Pruna 문서</a>에서 최적화 기법의 전체 범위를 살펴보세요.</p> <!--[1--><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 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><!---->설치<!----></span></h2><!--]--><!----> <p>다음 명령어로 Pruna를 설치합니다.</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> 복사됨</div><!----></button><!----></div> <pre class="language-bash "><!---->pip install pruna<!----></pre></div><!----> <!--[1--><h2 class="relative group"><a id="optimize-diffusers-models" 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="#optimize-diffusers-models"><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><!---->Diffusers 모델 최적화<!----></span></h2><!--]--><!----> <p>아래에 표시된 바와 같이 Diffusers 모델에는 다양한 최적화 알고리즘이 지원됩니다.</p> <div class="flex justify-center"><img src="https://huggingface.co/datasets/PrunaAI/documentation-images/resolve/main/diffusers/diffusers_combinations.png" alt="Overview of the supported optimization algorithms for diffusers models"/></div> <p>아래 예시에서는 factorizer, compiler, cacher 알고리즘의 조합을 사용하여 <a href="https://huggingface.co/black-forest-labs/FLUX.1-dev" rel="nofollow">black-forest-labs/FLUX.1-dev</a>를 최적화합니다. 이 조합은 추론 속도를 최대 4.2배까지 가속화하고, 피크 GPU 메모리 사용량을 34.7GB에서 28.0GB로 줄여주며, 출력 품질은 거의 동일하게 유지합니다.</p> <blockquote class="tip"><p>이 예제에서 사용된 최적화 기법에 대해 자세히 알아보려면 <a href="https://docs.pruna.ai/en/stable/docs_pruna/user_manual/configure.html" rel="nofollow">Pruna 최적화</a> 문서를 참조하세요.</p></blockquote> <div class="flex justify-center"><img src="https://huggingface.co/datasets/PrunaAI/documentation-images/resolve/main/diffusers/flux_combination.png" alt="Optimization techniques used for FLUX.1-dev showing the combination of factorizer, compiler, and cacher algorithms"/></div> <p>최적화 알고리즘을 사용하여 <code>SmashConfig</code>를 정의하는 것으로 시작합니다. 모델을 최적화하려면 파이프라인과 <code>SmashConfig</code>를 <code>smash</code>로 감싸고, 이후 추론을 위해 파이프라인을 정상적으로 사용하면 됩니다.</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> 복사됨</div><!----></button><!----></div> <pre class="language-python "><!----><span class="hljs-keyword">import</span> torch
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> FluxPipeline

<span class="hljs-keyword">from</span> pruna <span class="hljs-keyword">import</span> PrunaModel, SmashConfig, smash

<span class="hljs-comment"># load the model</span>
<span class="hljs-comment"># Try segmind/Segmind-Vega or black-forest-labs/FLUX.1-schnell with a small GPU memory</span>
pipe = FluxPipeline.from_pretrained(
    <span class="hljs-string">&quot;black-forest-labs/FLUX.1-dev&quot;</span>,
    dtype=torch.bfloat16
).to(<span class="hljs-string">&quot;cuda&quot;</span>)  <span class="hljs-comment"># or &quot;mps&quot;, &quot;xpu&quot;, &quot;cpu&quot;</span>

<span class="hljs-comment"># define the configuration</span>
smash_config = SmashConfig()
smash_config[<span class="hljs-string">&quot;factorizer&quot;</span>] = <span class="hljs-string">&quot;qkv_diffusers&quot;</span>
smash_config[<span class="hljs-string">&quot;compiler&quot;</span>] = <span class="hljs-string">&quot;torch_compile&quot;</span>
smash_config[<span class="hljs-string">&quot;torch_compile_target&quot;</span>] = <span class="hljs-string">&quot;module_list&quot;</span>
smash_config[<span class="hljs-string">&quot;cacher&quot;</span>] = <span class="hljs-string">&quot;fora&quot;</span>
smash_config[<span class="hljs-string">&quot;fora_interval&quot;</span>] = <span class="hljs-number">2</span>

<span class="hljs-comment"># for the best results in terms of speed you can add these configs</span>
<span class="hljs-comment"># however they will increase your warmup time from 1.5 min to 10 min</span>
<span class="hljs-comment"># smash_config[&quot;torch_compile_mode&quot;] = &quot;max-autotune-no-cudagraphs&quot;</span>
<span class="hljs-comment"># smash_config[&quot;quantizer&quot;] = &quot;torchao&quot;</span>
<span class="hljs-comment"># smash_config[&quot;torchao_quant_type&quot;] = &quot;fp8dq&quot;</span>
<span class="hljs-comment"># smash_config[&quot;torchao_excluded_modules&quot;] = &quot;norm+embedding&quot;</span>

<span class="hljs-comment"># optimize the model</span>
smashed_pipe = smash(pipe, smash_config)

<span class="hljs-comment"># run the model</span>
smashed_pipe(<span class="hljs-string">&quot;a knitted purple prune&quot;</span>).images[<span class="hljs-number">0</span>]<!----></pre></div><!----> <div class="flex justify-center"><img src="https://huggingface.co/datasets/PrunaAI/documentation-images/resolve/main/diffusers/flux_smashed_comparison.png"/></div> <p>최적화 후, Hugging Face Hub를 사용하여 최적화된 모델을 공유하고 로드할 수 있습니다.</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> 복사됨</div><!----></button><!----></div> <pre class="language-python "><!----><span class="hljs-comment"># save the model</span>
smashed_pipe.save_to_hub(<span class="hljs-string">&quot;&lt;username&gt;/FLUX.1-dev-smashed&quot;</span>)

<span class="hljs-comment"># load the model</span>
smashed_pipe = PrunaModel.from_hub(<span class="hljs-string">&quot;&lt;username&gt;/FLUX.1-dev-smashed&quot;</span>)<!----></pre></div><!----> <!--[1--><h2 class="relative group"><a id="evaluate-and-benchmark-diffusers-models" 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="#evaluate-and-benchmark-diffusers-models"><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><!---->Diffusers 모델 평가 및 벤치마킹<!----></span></h2><!--]--><!----> <p>Pruna는 최적화된 모델의 품질을 평가하기 위해 <a href="https://docs.pruna.ai/en/stable/docs_pruna/user_manual/evaluate.html" rel="nofollow">EvaluationAgent</a>를 제공합니다.</p> <p>총 시간과 처리량과 같이 관심 있는 metrics와 평가할 데이터셋을 정의할 수 있습니다. 모델을 정의하여 <code>EvaluationAgent</code>에 전달할 수 있습니다.</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">최적화된 모델</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">독립형 모델</div><!--]--></div> <div class="language-select"><div><p><code>EvaluationAgent</code>를 사용하여 최적화된 모델을 로드하고 평가한 다음, 이를 <code>Task</code>에 전달할 수 있습니다.</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> 복사됨</div><!----></button><!----></div> <pre class="language-python "><!----><span class="hljs-keyword">import</span> torch
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> FluxPipeline

<span class="hljs-keyword">from</span> pruna <span class="hljs-keyword">import</span> PrunaModel
<span class="hljs-keyword">from</span> pruna.data.pruna_datamodule <span class="hljs-keyword">import</span> PrunaDataModule
<span class="hljs-keyword">from</span> pruna.evaluation.evaluation_agent <span class="hljs-keyword">import</span> EvaluationAgent
<span class="hljs-keyword">from</span> pruna.evaluation.metrics <span class="hljs-keyword">import</span> (
    ThroughputMetric,
    TorchMetricWrapper,
    TotalTimeMetric,
)
<span class="hljs-keyword">from</span> pruna.evaluation.task <span class="hljs-keyword">import</span> Task

<span class="hljs-comment"># define the device</span>
device = <span class="hljs-string">&quot;cuda&quot;</span> <span class="hljs-keyword">if</span> torch.cuda.is_available() <span class="hljs-keyword">else</span> <span class="hljs-string">&quot;mps&quot;</span> <span class="hljs-keyword">if</span> torch.backends.mps.is_available() <span class="hljs-keyword">else</span> <span class="hljs-string">&quot;cpu&quot;</span>

<span class="hljs-comment"># load the model</span>
<span class="hljs-comment"># Try PrunaAI/Segmind-Vega-smashed or PrunaAI/FLUX.1-dev-smashed with a small GPU memory</span>
smashed_pipe = PrunaModel.from_hub(<span class="hljs-string">&quot;PrunaAI/FLUX.1-dev-smashed&quot;</span>)

<span class="hljs-comment"># Define the metrics</span>
metrics = [
    TotalTimeMetric(n_iterations=<span class="hljs-number">20</span>, n_warmup_iterations=<span class="hljs-number">5</span>),
    ThroughputMetric(n_iterations=<span class="hljs-number">20</span>, n_warmup_iterations=<span class="hljs-number">5</span>),
    TorchMetricWrapper(<span class="hljs-string">&quot;clip&quot;</span>),
]

<span class="hljs-comment"># Define the datamodule</span>
datamodule = PrunaDataModule.from_string(<span class="hljs-string">&quot;LAION256&quot;</span>)
datamodule.limit_datasets(<span class="hljs-number">10</span>)

<span class="hljs-comment"># Define the task and evaluation agent</span>
task = Task(metrics, datamodule=datamodule, device=device)
eval_agent = EvaluationAgent(task)

<span class="hljs-comment"># Evaluate smashed model and offload it to CPU</span>
smashed_pipe.move_to_device(device)
smashed_pipe_results = eval_agent.evaluate(smashed_pipe)
smashed_pipe.move_to_device(<span class="hljs-string">&quot;cpu&quot;</span>)<!----></pre></div><!----><!----></div><!----> <div hidden=""><p>최적화된 모델을 기본 모델과 비교하는 대신, 단독 <code>diffusers</code> 모델을 평가할 수도 있습니다. 이는 최적화 없이 모델의 성능을 평가하려는 경우에 유용합니다. <code>PrunaModel</code> 래퍼를 사용하고 해당 래퍼에서 <code>EvaluationAgent</code>를 실행하여 이를 수행할 수 있습니다.</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> 복사됨</div><!----></button><!----></div> <pre class="language-python "><!----><span class="hljs-keyword">import</span> torch
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> FluxPipeline

<span class="hljs-keyword">from</span> pruna <span class="hljs-keyword">import</span> PrunaModel

<span class="hljs-comment"># load the model</span>
<span class="hljs-comment"># Try PrunaAI/Segmind-Vega-smashed or PrunaAI/FLUX.1-dev-smashed with a small GPU memory</span>
pipe = FluxPipeline.from_pretrained(
    <span class="hljs-string">&quot;black-forest-labs/FLUX.1-dev&quot;</span>,
    dtype=torch.bfloat16
).to(<span class="hljs-string">&quot;cpu&quot;</span>)
wrapped_pipe = PrunaModel(model=pipe)<!----></pre></div><!----><!----></div><!----><!----></div><!----> <p>모델을 최적화하고 평가하는 방법을 확인하셨으니 이제 Pruna를 사용해 직접 모델을 최적화할 수 있습니다. 다행히도 시작을 돕기 위한 다양한 예제가 준비되어 있습니다.</p> <blockquote class="tip"><p>Flux 벤치마킹에 대한 자세한 내용은 <a href="https://huggingface.co/blog/PrunaAI/flux-fastest-image-generation-endpoint" rel="nofollow">Announcing FLUX-Juiced: The Fastest Image Generation Endpoint (2.6 times faster)!</a> 블로그 글과 <a href="https://huggingface.co/spaces/PrunaAI/InferBench" rel="nofollow">InferBench</a> Space를 참고하세요.</p></blockquote> <!--[1--><h2 class="relative group"><a id="reference" 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="#reference"><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><!---->참고<!----></span></h2><!--]--><!----> <ul><li><a href="https://github.com/PrunaAI/pruna" rel="nofollow">Pruna</a></li> <li><a href="https://docs.pruna.ai/en/stable/docs_pruna/user_manual/configure.html#configure-algorithms" rel="nofollow">Pruna 최적화</a></li> <li><a href="https://docs.pruna.ai/en/stable/docs_pruna/user_manual/evaluate.html" rel="nofollow">Pruna 평가</a></li> <li><a href="https://docs.pruna.ai/en/stable/docs_pruna/tutorials/index.html" rel="nofollow">Pruna 튜토리얼</a></li></ul> <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/en/optimization/pruna.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_1u9qigp = {
						base: "/docs/diffusers/main/en",
						assets: "/docs/diffusers/main/en"
					};
					const element = document.currentScript.parentElement;
					Promise.all([
						import("/docs/diffusers/main/en/_app/immutable/entry/start.CM-2NerT.js"),
						import("/docs/diffusers/main/en/_app/immutable/entry/app.DrYcqW7m.js")
					]).then(([kit, app]) => {
						kit.start(app, element, {
							node_ids: [0, 11],
							data: [null,null],
							form: null,
							error: null
						});
					});
				}
			</script>
		
Xet Storage Details

Size:: 31.3 kB
Xet hash:: 6dbac9c5beff01eb7347655fed54f149e01cd8d7ba36cc8d8ea0b23064215112
Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.