Hugging Face's logo

Buckets:
hf-doc-build
/
doc-dev

Files
xet
hf-doc-build/doc-dev / transformers /main /ko /custom_models.html
rtrm's picture
rtrm
download
raw
55.8 kB
<meta charset="utf-8" /><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;사용자 정의 모델 공유하기&quot;,&quot;local&quot;:&quot;sharing-custom-models&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;사용자 정의 구성 작성하기&quot;,&quot;local&quot;:&quot;writing-a-custom-configuration&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;사용자 정의 모델 작성하기&quot;,&quot;local&quot;:&quot;writing-a-custom-model&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Hub로 코드 업로드하기&quot;,&quot;local&quot;:&quot;sending-the-code-to-the-hub&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;사용자 정의 코드로 모델 사용하기&quot;,&quot;local&quot;:&quot;using-a-model-with-custom-code&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;사용자 정의 코드로 만든 모델을 auto 클래스로 등록하기&quot;,&quot;local&quot;:&quot;registering-a-model-with-custom-code-to-the-auto-classes&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2}],&quot;depth&quot;:1}">
 <link href="/docs/transformers/main/ko/_app/immutable/assets/0.e3b0c442.css" rel="modulepreload">
 <link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/entry/start.9aa88961.js">
 <link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/chunks/scheduler.9bc65507.js">
 <link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/chunks/singletons.9eec45c3.js">
 <link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/chunks/index.3b203c72.js">
 <link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/chunks/paths.566078f7.js">
 <link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/entry/app.84fb67c3.js">
 <link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/chunks/index.707bf1b6.js">
 <link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/nodes/0.1c99376b.js">
 <link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/chunks/each.e59479a4.js">
 <link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/nodes/14.1a1bfcfc.js">
 <link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/chunks/Tip.c2ecdbf4.js">
 <link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/chunks/CodeBlock.54a9f38d.js">
 <link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/chunks/EditOnGithub.922df6ba.js"><!-- HEAD_svelte-u9bgzb_START --><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;사용자 정의 모델 공유하기&quot;,&quot;local&quot;:&quot;sharing-custom-models&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;사용자 정의 구성 작성하기&quot;,&quot;local&quot;:&quot;writing-a-custom-configuration&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;사용자 정의 모델 작성하기&quot;,&quot;local&quot;:&quot;writing-a-custom-model&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Hub로 코드 업로드하기&quot;,&quot;local&quot;:&quot;sending-the-code-to-the-hub&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;사용자 정의 코드로 모델 사용하기&quot;,&quot;local&quot;:&quot;using-a-model-with-custom-code&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;사용자 정의 코드로 만든 모델을 auto 클래스로 등록하기&quot;,&quot;local&quot;:&quot;registering-a-model-with-custom-code-to-the-auto-classes&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2}],&quot;depth&quot;:1}"><!-- HEAD_svelte-u9bgzb_END --> <p></p> <h1 class="relative group"><a id="sharing-custom-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="#sharing-custom-models"><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>사용자 정의 모델 공유하기</span></h1> <p data-svelte-h="svelte-al07n">🤗 Transformers 라이브러리는 쉽게 확장할 수 있도록 설계되었습니다.
모든 모델은 추상화 없이 저장소의 지정된 하위 폴더에 완전히 코딩되어 있으므로, 손쉽게 모델링 파일을 복사하고 필요에 따라 조정할 수 있습니다.</p> <p data-svelte-h="svelte-1lxcxix">완전히 새로운 모델을 만드는 경우에는 처음부터 시작하는 것이 더 쉬울 수 있습니다.
이 튜토리얼에서는 Transformers 내에서 사용할 수 있도록 사용자 정의 모델과 구성을 작성하는 방법과
🤗 Transformers 라이브러리에 없는 경우에도 누구나 사용할 수 있도록 (의존성과 함께) 커뮤니티에 공유하는 방법을 배울 수 있습니다.</p> <p data-svelte-h="svelte-119rg2w"><a href="https://github.com/rwightman/pytorch-image-models" rel="nofollow">timm 라이브러리</a>의 ResNet 클래스를 <code>PreTrainedModel</code>로 래핑한 ResNet 모델을 예로 모든 것을 설명합니다.</p> <h2 class="relative group"><a id="writing-a-custom-configuration" 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="#writing-a-custom-configuration"><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>사용자 정의 구성 작성하기</span></h2> <p data-svelte-h="svelte-la43mp">모델에 들어가기 전에 먼저 구성을 작성해보도록 하겠습니다.
모델의 <code>configuration</code>은 모델을 만들기 위해 필요한 모든 중요한 것들을 포함하고 있는 객체입니다.
다음 섹션에서 볼 수 있듯이, 모델은 <code>config</code>를 사용해서만 초기화할 수 있기 때문에 완벽한 구성이 필요합니다.</p> <p data-svelte-h="svelte-1qc22zw">아래 예시에서는 ResNet 클래스의 인수(argument)를 조정해보겠습니다.
다른 구성은 가능한 ResNet 중 다른 유형을 제공합니다.
그런 다음 몇 가지 유효성을 확인한 후 해당 인수를 저장합니다.</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> transformers <span class="hljs-keyword">import</span> PretrainedConfig
<span class="hljs-keyword">from</span> typing <span class="hljs-keyword">import</span> <span class="hljs-type">List</span>
<span class="hljs-keyword">class</span> <span class="hljs-title class_">ResnetConfig</span>(<span class="hljs-title class_ inherited__">PretrainedConfig</span>):
 model_type = <span class="hljs-string">&quot;resnet&quot;</span>
<span class="hljs-keyword">def</span> <span class="hljs-title function_">__init__</span>(<span class="hljs-params">
 self,
 block_type=<span class="hljs-string">&quot;bottleneck&quot;</span>,
 layers: <span class="hljs-type">List</span>[<span class="hljs-built_in">int</span>] = [<span class="hljs-number">3</span>, <span class="hljs-number">4</span>, <span class="hljs-number">6</span>, <span class="hljs-number">3</span>],
 num_classes: <span class="hljs-built_in">int</span> = <span class="hljs-number">1000</span>,
 input_channels: <span class="hljs-built_in">int</span> = <span class="hljs-number">3</span>,
 cardinality: <span class="hljs-built_in">int</span> = <span class="hljs-number">1</span>,
 base_width: <span class="hljs-built_in">int</span> = <span class="hljs-number">64</span>,
 stem_width: <span class="hljs-built_in">int</span> = <span class="hljs-number">64</span>,
 stem_type: <span class="hljs-built_in">str</span> = <span class="hljs-string">&quot;&quot;</span>,
 avg_down: <span class="hljs-built_in">bool</span> = <span class="hljs-literal">False</span>,
 **kwargs,
 </span>):
 <span class="hljs-keyword">if</span> block_type <span class="hljs-keyword">not</span> <span class="hljs-keyword">in</span> [<span class="hljs-string">&quot;basic&quot;</span>, <span class="hljs-string">&quot;bottleneck&quot;</span>]:
 <span class="hljs-keyword">raise</span> ValueError(<span class="hljs-string">f&quot;`block_type` must be &#x27;basic&#x27; or bottleneck&#x27;, got <span class="hljs-subst">{block_type}</span>.&quot;</span>)
 <span class="hljs-keyword">if</span> stem_type <span class="hljs-keyword">not</span> <span class="hljs-keyword">in</span> [<span class="hljs-string">&quot;&quot;</span>, <span class="hljs-string">&quot;deep&quot;</span>, <span class="hljs-string">&quot;deep-tiered&quot;</span>]:
 <span class="hljs-keyword">raise</span> ValueError(<span class="hljs-string">f&quot;`stem_type` must be &#x27;&#x27;, &#x27;deep&#x27; or &#x27;deep-tiered&#x27;, got <span class="hljs-subst">{stem_type}</span>.&quot;</span>)
self.block_type = block_type
 self.layers = layers
 self.num_classes = num_classes
 self.input_channels = input_channels
 self.cardinality = cardinality
 self.base_width = base_width
 self.stem_width = stem_width
 self.stem_type = stem_type
 self.avg_down = avg_down
 <span class="hljs-built_in">super</span>().__init__(**kwargs)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-nxq6mo">사용자 정의 <code>configuration</code>을 작성할 때 기억해야 할 세 가지 중요한 사항은 다음과 같습니다:</p> <ul data-svelte-h="svelte-1tfxli8"><li><code>PretrainedConfig</code>을 상속해야 합니다.</li> <li><code>PretrainedConfig</code>의 <code>__init__</code>은 모든 kwargs를 허용해야 하고,</li> <li>이러한 <code>kwargs</code>는 상위 클래스 <code>__init__</code>에 전달되어야 합니다.</li></ul> <p data-svelte-h="svelte-x5ckqo">상속은 🤗 Transformers 라이브러리에서 모든 기능을 가져오는 것입니다.
이러한 점으로부터 비롯되는 두 가지 제약 조건은 <code>PretrainedConfig</code>에 설정하는 것보다 더 많은 필드가 있습니다.
<code>from_pretrained</code> 메서드로 구성을 다시 로드할 때 해당 필드는 구성에서 수락한 후 상위 클래스로 보내야 합니다.</p> <p data-svelte-h="svelte-1kaub1g">모델을 auto 클래스에 등록하지 않는 한, <code>configuration</code>에서 <code>model_type</code>을 정의(여기서 <code>model_type=&quot;resnet&quot;</code>)하는 것은 필수 사항이 아닙니다 (마지막 섹션 참조).</p> <p data-svelte-h="svelte-ovl0zx">이렇게 하면 라이브러리의 다른 모델 구성과 마찬가지로 구성을 쉽게 만들고 저장할 수 있습니다.
다음은 resnet50d 구성을 생성하고 저장하는 방법입니다:</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 -->resnet50d_config = ResnetConfig(block_type=<span class="hljs-string">&quot;bottleneck&quot;</span>, stem_width=<span class="hljs-number">32</span>, stem_type=<span class="hljs-string">&quot;deep&quot;</span>, avg_down=<span class="hljs-literal">True</span>)
resnet50d_config.save_pretrained(<span class="hljs-string">&quot;custom-resnet&quot;</span>)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-d1xscf">이렇게 하면 <code>custom-resnet</code> 폴더 안에 <code>config.json</code>이라는 파일이 저장됩니다.
그런 다음 <code>from_pretrained</code> 메서드를 사용하여 구성을 다시 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 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 -->resnet50d_config = ResnetConfig.from_pretrained(<span class="hljs-string">&quot;custom-resnet&quot;</span>)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-g1m52h">구성을 Hub에 직접 업로드하기 위해 <code>PretrainedConfig</code> 클래스의 <code>push_to_hub()</code>와 같은 다른 메서드를 사용할 수 있습니다.</p> <h2 class="relative group"><a id="writing-a-custom-model" 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="#writing-a-custom-model"><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>사용자 정의 모델 작성하기</span></h2> <p data-svelte-h="svelte-6nuh70">이제 ResNet 구성이 있으므로 모델을 작성할 수 있습니다.
실제로는 두 개를 작성할 것입니다. 하나는 이미지 배치에서 hidden features를 추출하는 것(<code>BertModel</code>과 같이), 다른 하나는 이미지 분류에 적합한 것입니다(<code>BertForSequenceClassification</code>과 같이).</p> <p data-svelte-h="svelte-1ufn8h6">이전에 언급했듯이 이 예제에서는 단순하게 하기 위해 모델의 느슨한 래퍼(loose wrapper)만 작성할 것입니다.
이 클래스를 작성하기 전에 블록 유형과 실제 블록 클래스 간의 매핑 작업만 하면 됩니다.
그런 다음 <code>ResNet</code> 클래스로 전달되어 <code>configuration</code>을 통해 모델이 선언됩니다:</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> transformers <span class="hljs-keyword">import</span> PreTrainedModel
<span class="hljs-keyword">from</span> timm.models.resnet <span class="hljs-keyword">import</span> BasicBlock, Bottleneck, ResNet
<span class="hljs-keyword">from</span> .configuration_resnet <span class="hljs-keyword">import</span> ResnetConfig
BLOCK_MAPPING = {<span class="hljs-string">&quot;basic&quot;</span>: BasicBlock, <span class="hljs-string">&quot;bottleneck&quot;</span>: Bottleneck}
<span class="hljs-keyword">class</span> <span class="hljs-title class_">ResnetModel</span>(<span class="hljs-title class_ inherited__">PreTrainedModel</span>):
 config_class = ResnetConfig
<span class="hljs-keyword">def</span> <span class="hljs-title function_">__init__</span>(<span class="hljs-params">self, config</span>):
 <span class="hljs-built_in">super</span>().__init__(config)
 block_layer = BLOCK_MAPPING[config.block_type]
 self.model = ResNet(
 block_layer,
 config.layers,
 num_classes=config.num_classes,
 in_chans=config.input_channels,
 cardinality=config.cardinality,
 base_width=config.base_width,
 stem_width=config.stem_width,
 stem_type=config.stem_type,
 avg_down=config.avg_down,
 )
<span class="hljs-keyword">def</span> <span class="hljs-title function_">forward</span>(<span class="hljs-params">self, tensor</span>):
 <span class="hljs-keyword">return</span> self.model.forward_features(tensor)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-10szka3">이미지 분류 모델을 만들기 위해서는 forward 메소드만 변경하면 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 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">class</span> <span class="hljs-title class_">ResnetModelForImageClassification</span>(<span class="hljs-title class_ inherited__">PreTrainedModel</span>):
 config_class = ResnetConfig
<span class="hljs-keyword">def</span> <span class="hljs-title function_">__init__</span>(<span class="hljs-params">self, config</span>):
 <span class="hljs-built_in">super</span>().__init__(config)
 block_layer = BLOCK_MAPPING[config.block_type]
 self.model = ResNet(
 block_layer,
 config.layers,
 num_classes=config.num_classes,
 in_chans=config.input_channels,
 cardinality=config.cardinality,
 base_width=config.base_width,
 stem_width=config.stem_width,
 stem_type=config.stem_type,
 avg_down=config.avg_down,
 )
<span class="hljs-keyword">def</span> <span class="hljs-title function_">forward</span>(<span class="hljs-params">self, tensor, labels=<span class="hljs-literal">None</span></span>):
 logits = self.model(tensor)
 <span class="hljs-keyword">if</span> labels <span class="hljs-keyword">is</span> <span class="hljs-keyword">not</span> <span class="hljs-literal">None</span>:
 loss = torch.nn.functional.cross_entropy(logits, labels)
 <span class="hljs-keyword">return</span> {<span class="hljs-string">&quot;loss&quot;</span>: loss, <span class="hljs-string">&quot;logits&quot;</span>: logits}
 <span class="hljs-keyword">return</span> {<span class="hljs-string">&quot;logits&quot;</span>: logits}<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1g3ya33">두 경우 모두 <code>PreTrainedModel</code>를 상속받고, <code>config</code>를 통해 상위 클래스 초기화를 호출하다는 점을 기억하세요 (일반적인 <code>torch.nn.Module</code>을 작성할 때와 비슷함).
모델을 auto 클래스에 등록하고 싶은 경우에는 <code>config_class</code>를 설정하는 부분이 필수입니다 (마지막 섹션 참조).</p> <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"><p data-svelte-h="svelte-hf6zh1">라이브러리에 존재하는 모델과 굉장히 유사하다면, 모델을 생성할 때 구성을 참조해 재사용할 수 있습니다.</p></div> <p data-svelte-h="svelte-1f99hcg">원하는 것을 모델이 반환하도록 할 수 있지만, <code>ResnetModelForImageClassification</code>에서 했던 것 처럼
레이블을 통과시켰을 때 손실과 함께 사전 형태로 반환하는 것이 <code>Trainer</code> 클래스 내에서 직접 모델을 사용하기에 유용합니다.
자신만의 학습 루프 또는 다른 학습 라이브러리를 사용할 계획이라면 다른 출력 형식을 사용해도 좋습니다.</p> <p data-svelte-h="svelte-68rs2x">이제 모델 클래스가 있으므로 하나 생성해 보겠습니다:</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 -->resnet50d = ResnetModelForImageClassification(resnet50d_config)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-10d07s4">다시 말하지만, <code>save_pretrained()</code>또는 <code>push_to_hub()</code>처럼 <code>PreTrainedModel</code>에 속하는 모든 메소드를 사용할 수 있습니다.
다음 섹션에서 두 번째 메소드를 사용해 모델 코드와 모델 가중치를 업로드하는 방법을 살펴보겠습니다.
먼저, 모델 내부에 사전 훈련된 가중치를 로드해 보겠습니다.</p> <p data-svelte-h="svelte-1319e67">이 예제를 활용할 때는, 사용자 정의 모델을 자신만의 데이터로 학습시킬 것입니다.
이 튜토리얼에서는 빠르게 진행하기 위해 사전 훈련된 resnet50d를 사용하겠습니다.
아래 모델은 resnet50d의 래퍼이기 때문에, 가중치를 쉽게 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 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> timm
pretrained_model = timm.create_model(<span class="hljs-string">&quot;resnet50d&quot;</span>, pretrained=<span class="hljs-literal">True</span>)
resnet50d.model.load_state_dict(pretrained_model.state_dict())<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-rtsgc0">이제 <code>save_pretrained()</code> 또는 <code>push_to_hub()</code>를 사용할 때 모델 코드가 저장되는지 확인해봅시다.</p> <h2 class="relative group"><a id="sending-the-code-to-the-hub" 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="#sending-the-code-to-the-hub"><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>Hub로 코드 업로드하기</span></h2> <div class="course-tip course-tip-orange bg-gradient-to-br dark:bg-gradient-to-r before:border-orange-500 dark:before:border-orange-800 from-orange-50 dark:from-gray-900 to-white dark:to-gray-950 border border-orange-50 text-orange-700 dark:text-gray-400"><p data-svelte-h="svelte-11fial7">이 API는 실험적이며 다음 릴리스에서 약간의 변경 사항이 있을 수 있습니다.</p></div> <p data-svelte-h="svelte-q87di4">먼저 모델이 <code>.py</code> 파일에 완전히 정의되어 있는지 확인하세요.
모든 파일이 동일한 작업 경로에 있기 때문에 상대경로 임포트(relative import)에 의존할 수 있습니다 (transformers에서는 이 기능에 대한 하위 모듈을 지원하지 않습니다).
이 예시에서는 작업 경로 안의 <code>resnet_model</code>에서 <code>modeling_resnet.py</code> 파일과 <code>configuration_resnet.py</code> 파일을 정의합니다.
구성 파일에는 <code>ResnetConfig</code>에 대한 코드가 있고 모델링 파일에는 <code>ResnetModel</code> 및 <code>ResnetModelForImageClassification</code>에 대한 코드가 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 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 -->.
└── resnet_model
 ├── __init__.<span class="hljs-keyword">py</span>
 ├── configuration_resnet.<span class="hljs-keyword">py</span>
 └── modeling_resnet.<span class="hljs-keyword">py</span><!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-fw3tvd">Python이 <code>resnet_model</code>을 모듈로 사용할 수 있도록 감지하는 목적이기 때문에 <code>__init__.py</code>는 비어 있을 수 있습니다.</p> <div class="course-tip course-tip-orange bg-gradient-to-br dark:bg-gradient-to-r before:border-orange-500 dark:before:border-orange-800 from-orange-50 dark:from-gray-900 to-white dark:to-gray-950 border border-orange-50 text-orange-700 dark:text-gray-400"><p data-svelte-h="svelte-1igyulw">라이브러리에서 모델링 파일을 복사하는 경우,
모든 파일 상단에 있는 상대 경로 임포트(relative import) 부분을 <code>transformers</code> 패키지에서 임포트 하도록 변경해야 합니다.</p></div> <p data-svelte-h="svelte-19tqquy">기존 구성이나 모델을 재사용(또는 서브 클래스화)할 수 있습니다.</p> <p data-svelte-h="svelte-1tgl3y3">커뮤니티에 모델을 공유하기 위해서는 다음 단계를 따라야 합니다:
먼저, 새로 만든 파일에 ResNet 모델과 구성을 임포트합니다:</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> resnet_model.configuration_resnet <span class="hljs-keyword">import</span> ResnetConfig
<span class="hljs-keyword">from</span> resnet_model.modeling_resnet <span class="hljs-keyword">import</span> ResnetModel, ResnetModelForImageClassification<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-85qz6h">다음으로 <code>save_pretrained</code> 메소드를 사용해 해당 객체의 코드 파일을 복사하고,
복사한 파일을 Auto 클래스로 등록하고(모델인 경우) 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 -->ResnetConfig.register_for_auto_class()
ResnetModel.register_for_auto_class(<span class="hljs-string">&quot;AutoModel&quot;</span>)
ResnetModelForImageClassification.register_for_auto_class(<span class="hljs-string">&quot;AutoModelForImageClassification&quot;</span>)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-urqb2q"><code>configuration</code>에 대한 auto 클래스를 지정할 필요는 없지만(<code>configuration</code> 관련 auto 클래스는 AutoConfig 클래스 하나만 있음), 모델의 경우에는 지정해야 합니다.
사용자 지정 모델은 다양한 작업에 적합할 수 있으므로, 모델에 맞는 auto 클래스를 지정해야 합니다.</p> <p data-svelte-h="svelte-th8o9z">다음으로, 이전에 작업했던 것과 마찬가지로 구성과 모델을 작성합니다:</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 -->resnet50d_config = ResnetConfig(block_type=<span class="hljs-string">&quot;bottleneck&quot;</span>, stem_width=<span class="hljs-number">32</span>, stem_type=<span class="hljs-string">&quot;deep&quot;</span>, avg_down=<span class="hljs-literal">True</span>)
resnet50d = ResnetModelForImageClassification(resnet50d_config)
pretrained_model = timm.create_model(<span class="hljs-string">&quot;resnet50d&quot;</span>, pretrained=<span class="hljs-literal">True</span>)
resnet50d.model.load_state_dict(pretrained_model.state_dict())<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1xhzef1">이제 모델을 Hub로 업로드하기 위해 로그인 상태인지 확인하세요.
터미널에서 다음 코드를 실행해 확인할 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 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 -->huggingface-cli login<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-4bt266">주피터 노트북의 경우에는 다음과 같습니다:</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> huggingface_hub <span class="hljs-keyword">import</span> notebook_login
notebook_login()<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-xc5sk0">그런 다음 이렇게 자신의 네임스페이스(또는 자신이 속한 조직)에 업로드할 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 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 -->resnet50d.push_to_hub(<span class="hljs-string">&quot;custom-resnet50d&quot;</span>)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-dnw5aw">On top of the modeling weights and the configuration in json format, this also copied the modeling and
configuration <code>.py</code> files in the folder <code>custom-resnet50d</code> and uploaded the result to the Hub. You can check the result
in this <a href="https://huggingface.co/sgugger/custom-resnet50d" rel="nofollow">model repo</a>.
json 형식의 모델링 가중치와 구성 외에도 <code>custom-resnet50d</code> 폴더 안의 모델링과 구성 <code>.py</code> 파일을 복사하해 Hub에 업로드합니다.
<a href="https://huggingface.co/sgugger/custom-resnet50d" rel="nofollow">모델 저장소</a>에서 결과를 확인할 수 있습니다.</p> <p data-svelte-h="svelte-2rsbbb"><a href="model_sharing">sharing tutorial</a> 문서의 <code>push_to_hub</code> 메소드에서 자세한 내용을 확인할 수 있습니다.</p> <h2 class="relative group"><a id="using-a-model-with-custom-code" 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="#using-a-model-with-custom-code"><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>사용자 정의 코드로 모델 사용하기</span></h2> <p data-svelte-h="svelte-1m1cooj">auto 클래스와 <code>from_pretrained</code> 메소드를 사용하여 사용자 지정 코드 파일과 함께 모든 구성, 모델, 토크나이저를 사용할 수 있습니다.
Hub에 업로드된 모든 파일 및 코드는 멜웨어가 있는지 검사되지만 (자세한 내용은 <a href="https://huggingface.co/docs/hub/security#malware-scanning" rel="nofollow">Hub 보안</a> 설명 참조),
자신의 컴퓨터에서 모델 코드와 작성자가 악성 코드를 실행하지 않는지 확인해야 합니다.
사용자 정의 코드로 모델을 사용하려면 <code>trust_remote_code=True</code>로 설정하세요:</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> transformers <span class="hljs-keyword">import</span> AutoModelForImageClassification
model = AutoModelForImageClassification.from_pretrained(<span class="hljs-string">&quot;sgugger/custom-resnet50d&quot;</span>, trust_remote_code=<span class="hljs-literal">True</span>)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-18oh14g">모델 작성자가 악의적으로 코드를 업데이트하지 않았다는 점을 확인하기 위해, 커밋 해시(commit hash)를 <code>revision</code>으로 전달하는 것도 강력히 권장됩니다 (모델 작성자를 완전히 신뢰하지 않는 경우).</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 -->commit_hash = <span class="hljs-string">&quot;ed94a7c6247d8aedce4647f00f20de6875b5b292&quot;</span>
model = AutoModelForImageClassification.from_pretrained(
 <span class="hljs-string">&quot;sgugger/custom-resnet50d&quot;</span>, trust_remote_code=<span class="hljs-literal">True</span>, revision=commit_hash
)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-eerkzr">Hub에서 모델 저장소의 커밋 기록을 찾아볼 때, 모든 커밋의 커밋 해시를 쉽게 복사할 수 있는 버튼이 있습니다.</p> <h2 class="relative group"><a id="registering-a-model-with-custom-code-to-the-auto-classes" 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="#registering-a-model-with-custom-code-to-the-auto-classes"><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>사용자 정의 코드로 만든 모델을 auto 클래스로 등록하기</span></h2> <p data-svelte-h="svelte-1w0yfzt">🤗 Transformers를 상속하는 라이브러리를 작성하는 경우 사용자 정의 모델을 auto 클래스에 추가할 수 있습니다.
사용자 정의 모델을 사용하기 위해 해당 라이브러리를 임포트해야 하기 때문에, 이는 Hub로 코드를 업로드하는 것과 다릅니다 (Hub에서 자동적으로 모델 코드를 다운로드 하는 것과 반대).</p> <p data-svelte-h="svelte-xh4cqs">구성에 기존 모델 유형과 다른 <code>model_type</code> 속성이 있고 모델 클래스에 올바른 <code>config_class</code> 속성이 있는 한,
다음과 같이 auto 클래스에 추가할 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 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> transformers <span class="hljs-keyword">import</span> AutoConfig, AutoModel, AutoModelForImageClassification
AutoConfig.register(<span class="hljs-string">&quot;resnet&quot;</span>, ResnetConfig)
AutoModel.register(ResnetConfig, ResnetModel)
AutoModelForImageClassification.register(ResnetConfig, ResnetModelForImageClassification)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-9j79lz">사용자 정의 구성을 <code>AutoConfig</code>에 등록할 때 사용되는 첫 번째 인수는 사용자 정의 구성의 <code>model_type</code>과 일치해야 합니다.
또한, 사용자 정의 모델을 auto 클래스에 등록할 때 사용되는 첫 번째 인수는 해당 모델의 <code>config_class</code>와 일치해야 합니다.</p> <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/ko/custom_models.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_1hrx8 = {
 assets: "/docs/transformers/main/ko",
 base: "/docs/transformers/main/ko",
 env: {}
 };
 const element = document.currentScript.parentElement;
 const data = [null,null];
 Promise.all([
 import("/docs/transformers/main/ko/_app/immutable/entry/start.9aa88961.js"),
 import("/docs/transformers/main/ko/_app/immutable/entry/app.84fb67c3.js")
 ]).then(([kit, app]) => {
 kit.start(app, element, {
 node_ids: [0, 14],
 data,
 form: null,
 error: null
 });
 });
 }
 </script>

Xet Storage Details

Size:
55.8 kB
·
Xet hash:
a1b796609db13727a40db3970a1adf67450d5ee093f622b37831eae297fdf8a1

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