Buckets:
| <meta charset="utf-8" /><meta name="hf:doc:metadata" content="{"title":"문제 해결","local":"troubleshoot","sections":[{"title":"방화벽 환경","local":"firewalled-environments","sections":[],"depth":2},{"title":"CUDA 메모리 부족(CUDA out of memory)","local":"cuda-out-of-memory","sections":[],"depth":2},{"title":"저장된 TensorFlow 모델을 가져올 수 없습니다(Unable to load a saved TensorFlow model)","local":"unable-to-load-a-saved-uensorFlow-model","sections":[],"depth":2},{"title":"ImportError","local":"importerror","sections":[],"depth":2},{"title":"CUDA error: device-side assert triggered","local":"cuda-error-deviceside-assert-triggered","sections":[],"depth":2},{"title":"패딩 토큰이 마스킹되지 않은 경우 잘못된 출력(Incorrect output when padding tokens aren’t masked)","local":"incorrect-output-when-padding-tokens-arent-masked","sections":[],"depth":2},{"title":"ValueError: 이 유형의 AutoModel에 대해 인식할 수 없는 XYZ 구성 클래스(ValueError: Unrecognized configuration class XYZ for this kind of AutoModel)","local":"valueerror-unrecognized-configuration-class-xyz-for-this-kind-of-automodel","sections":[],"depth":2}],"depth":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/96.985e79ac.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/Youtube.e1129c6f.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="{"title":"문제 해결","local":"troubleshoot","sections":[{"title":"방화벽 환경","local":"firewalled-environments","sections":[],"depth":2},{"title":"CUDA 메모리 부족(CUDA out of memory)","local":"cuda-out-of-memory","sections":[],"depth":2},{"title":"저장된 TensorFlow 모델을 가져올 수 없습니다(Unable to load a saved TensorFlow model)","local":"unable-to-load-a-saved-uensorFlow-model","sections":[],"depth":2},{"title":"ImportError","local":"importerror","sections":[],"depth":2},{"title":"CUDA error: device-side assert triggered","local":"cuda-error-deviceside-assert-triggered","sections":[],"depth":2},{"title":"패딩 토큰이 마스킹되지 않은 경우 잘못된 출력(Incorrect output when padding tokens aren’t masked)","local":"incorrect-output-when-padding-tokens-arent-masked","sections":[],"depth":2},{"title":"ValueError: 이 유형의 AutoModel에 대해 인식할 수 없는 XYZ 구성 클래스(ValueError: Unrecognized configuration class XYZ for this kind of AutoModel)","local":"valueerror-unrecognized-configuration-class-xyz-for-this-kind-of-automodel","sections":[],"depth":2}],"depth":1}"><!-- HEAD_svelte-u9bgzb_END --> <p></p> <h1 class="relative group"><a id="troubleshoot" 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="#troubleshoot"><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>문제 해결</span></h1> <p data-svelte-h="svelte-lsu3nl">때때로 오류가 발생할 수 있지만, 저희가 도와드리겠습니다! 이 가이드는 현재까지 확인된 가장 일반적인 문제 몇 가지와 그것들을 해결하는 방법에 대해 다룹니다. 그러나 이 가이드는 모든 🤗 Transformers 문제를 포괄적으로 다루고 있지 않습니다. 문제 해결에 더 많은 도움을 받으려면 다음을 시도해보세요:</p> <iframe class="w-full xl:w-4/6 h-80" src="https://www.youtube-nocookie.com/embed/S2EEG3JIt2A" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" allowfullscreen></iframe> <ol data-svelte-h="svelte-ucur6z"><li><a href="https://discuss.huggingface.co/" rel="nofollow">포럼</a>에서 도움을 요청하세요. <a href="https://discuss.huggingface.co/c/beginners/5" rel="nofollow">Beginners</a> 또는 <a href="https://discuss.huggingface.co/c/transformers/9" rel="nofollow">🤗 Transformers</a>와 같은 특정 카테고리에 질문을 게시할 수 있습니다. 재현 가능한 코드와 함께 잘 서술된 포럼 게시물을 작성하여 여러분의 문제가 해결될 가능성을 극대화하세요!</li></ol> <iframe class="w-full xl:w-4/6 h-80" src="https://www.youtube-nocookie.com/embed/_PAli-V4wj0" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" allowfullscreen></iframe> <ol start="2" data-svelte-h="svelte-uo4aew"><li><p>라이브러리와 관련된 버그이면 🤗 Transformers 저장소에서 <a href="https://github.com/huggingface/transformers/issues/new/choose" rel="nofollow">이슈</a>를 생성하세요. 버그에 대해 설명하는 정보를 가능한 많이 포함하려고 노력하여, 무엇이 잘못 되었는지와 어떻게 수정할 수 있는지 더 잘 파악할 수 있도록 도와주세요.</p></li> <li><p>이전 버전의 🤗 Transformers을 사용하는 경우 중요한 변경 사항이 버전 사이에 도입되었기 때문에 <a href="migration">마이그레이션</a> 가이드를 확인하세요.</p></li></ol> <p data-svelte-h="svelte-1uhar05">문제 해결 및 도움 매뉴얼에 대한 자세한 내용은 Hugging Face 강좌의 <a href="https://huggingface.co/course/chapter8/1?fw=pt" rel="nofollow">8장</a>을 참조하세요.</p> <h2 class="relative group"><a id="firewalled-environments" 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="#firewalled-environments"><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>방화벽 환경</span></h2> <p data-svelte-h="svelte-12ucoaj">클라우드 및 내부망(intranet) 설정의 일부 GPU 인스턴스는 외부 연결에 대한 방화벽으로 차단되어 연결 오류가 발생할 수 있습니다. 스크립트가 모델 가중치나 데이터를 다운로드하려고 할 때, 다운로드가 중단되고 다음 메시지와 함께 시간 초과됩니다:</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 -->ValueError: Connection error, <span class="hljs-built_in">and</span> we cannot <span class="hljs-keyword">find</span> the requested <span class="hljs-keyword">files</span> in the cached path. | |
| Please <span class="hljs-keyword">try</span> again <span class="hljs-built_in">or</span> <span class="hljs-keyword">make</span> sure your Internet connection <span class="hljs-keyword">is</span> <span class="hljs-keyword">on</span>.<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-11zq317">이 경우에는 연결 오류를 피하기 위해 🤗 Transformers를 <a href="installation#offline-mode">오프라인 모드</a>로 실행해야 합니다.</p> <h2 class="relative group"><a id="cuda-out-of-memory" 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="#cuda-out-of-memory"><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>CUDA 메모리 부족(CUDA out of memory)</span></h2> <p data-svelte-h="svelte-1c8hb26">수백만 개의 매개변수로 대규모 모델을 훈련하는 것은 적절한 하드웨어 없이 어려울 수 있습니다. GPU 메모리가 부족한 경우 발생할 수 있는 일반적인 오류는 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-attribute">CUDA</span> out of memory. Tried to allocate <span class="hljs-number">256</span>.<span class="hljs-number">00</span> MiB (GPU <span class="hljs-number">0</span>; <span class="hljs-number">11</span>.<span class="hljs-number">17</span> GiB total capacity; <span class="hljs-number">9</span>.<span class="hljs-number">70</span> GiB already allocated; <span class="hljs-number">179</span>.<span class="hljs-number">81</span> MiB free; <span class="hljs-number">9</span>.<span class="hljs-number">85</span> GiB reserved in total by PyTorch)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-15g2fcx">다음은 메모리 사용을 줄이기 위해 시도해 볼 수 있는 몇 가지 잠재적인 해결책입니다:</p> <ul data-svelte-h="svelte-meem4b"><li><code>TrainingArguments</code>의 <a href="main_classes/trainer#transformers.TrainingArguments.per_device_train_batch_size"><code>per_device_train_batch_size</code></a> 값을 줄이세요.</li> <li><code>TrainingArguments</code>의 <a href="main_classes/trainer#transformers.TrainingArguments.gradient_accumulation_steps"><code>gradient_accumulation_steps</code></a>은 전체 배치 크기를 효과적으로 늘리세요.</li></ul> <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-74wrmh">메모리 절약 기술에 대한 자세한 내용은 성능 <a href="performance">가이드</a>를 참조하세요.</p></div> <h2 class="relative group"><a id="unable-to-load-a-saved-uensorFlow-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="#unable-to-load-a-saved-uensorFlow-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>저장된 TensorFlow 모델을 가져올 수 없습니다(Unable to load a saved TensorFlow model)</span></h2> <p data-svelte-h="svelte-1v88cop">TensorFlow의 <a href="https://www.tensorflow.org/tutorials/keras/save_and_load#save_the_entire_model" rel="nofollow">model.save</a> 메소드는 아키텍처, 가중치, 훈련 구성 등 전체 모델을 단일 파일에 저장합니다. 그러나 모델 파일을 다시 가져올 때 🤗 Transformers는 모델 파일에 있는 모든 TensorFlow 관련 객체를 가져오지 않을 수 있기 때문에 오류가 발생할 수 있습니다. TensorFlow 모델 저장 및 가져오기 문제를 피하려면 다음을 권장합니다:</p> <ul data-svelte-h="svelte-colwgc"><li>모델 가중치를 <code>h5</code> 파일 확장자로 <a href="https://www.tensorflow.org/tutorials/keras/save_and_load#save_the_entire_model" rel="nofollow"><code>model.save_weights</code></a>로 저장한 다음 <code>from_pretrained()</code>로 모델을 다시 가져옵니다:</li></ul> <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-meta">>>> </span><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> TFPreTrainedModel | |
| <span class="hljs-meta">>>> </span><span class="hljs-keyword">from</span> tensorflow <span class="hljs-keyword">import</span> keras | |
| <span class="hljs-meta">>>> </span>model.save_weights(<span class="hljs-string">"some_folder/tf_model.h5"</span>) | |
| <span class="hljs-meta">>>> </span>model = TFPreTrainedModel.from_pretrained(<span class="hljs-string">"some_folder"</span>)<!-- HTML_TAG_END --></pre></div> <ul data-svelte-h="svelte-15oz86s"><li>모델을 <code>~TFPretrainedModel.save_pretrained</code>로 저장하고 <code>from_pretrained()</code>로 다시 가져옵니다:</li></ul> <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-meta">>>> </span><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> TFPreTrainedModel | |
| <span class="hljs-meta">>>> </span>model.save_pretrained(<span class="hljs-string">"path_to/model"</span>) | |
| <span class="hljs-meta">>>> </span>model = TFPreTrainedModel.from_pretrained(<span class="hljs-string">"path_to/model"</span>)<!-- HTML_TAG_END --></pre></div> <h2 class="relative group"><a id="importerror" 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="#importerror"><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>ImportError</span></h2> <p data-svelte-h="svelte-1ujp2u8">특히 최신 모델인 경우 만날 수 있는 다른 일반적인 오류는 <code>ImportError</code>입니다:</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 -->ImportError: cannot <span class="hljs-keyword">import</span> <span class="hljs-type">name</span> <span class="hljs-string">'ImageGPTImageProcessor'</span> <span class="hljs-keyword">from</span> <span class="hljs-string">'transformers'</span> (<span class="hljs-type">unknown</span> <span class="hljs-keyword">location</span>)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-m1x930">이러한 오류 유형의 경우 최신 모델에 액세스할 수 있도록 최신 버전의 🤗 Transformers가 설치되어 있는지 확인하세요:</p> <div class="code-block relative"><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->pip install transformers --upgrade<!-- HTML_TAG_END --></pre></div> <h2 class="relative group"><a id="cuda-error-deviceside-assert-triggered" 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="#cuda-error-deviceside-assert-triggered"><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>CUDA error: device-side assert triggered</span></h2> <p data-svelte-h="svelte-1tlsczl">때때로 장치 코드 오류에 대한 일반적인 CUDA 오류가 발생할 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 -->RuntimeError: CUDA <span class="hljs-literal">error</span>: device-<span class="hljs-literal">side</span> <span class="hljs-keyword">assert</span> triggered<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-17874yl">더 자세한 오류 메시지를 얻으려면 우선 코드를 CPU에서 실행합니다. 다음 환경 변수를 코드의 시작 부분에 추가하여 CPU로 전환하세요:</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-meta">>>> </span><span class="hljs-keyword">import</span> os | |
| <span class="hljs-meta">>>> </span>os.environ[<span class="hljs-string">"CUDA_VISIBLE_DEVICES"</span>] = <span class="hljs-string">""</span><!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-eyhzxh">또 다른 옵션은 GPU에서 더 나은 역추적(traceback)을 얻는 것입니다. 다음 환경 변수를 코드의 시작 부분에 추가하여 역추적이 오류가 발생한 소스를 가리키도록 하세요:</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-meta">>>> </span><span class="hljs-keyword">import</span> os | |
| <span class="hljs-meta">>>> </span>os.environ[<span class="hljs-string">"CUDA_LAUNCH_BLOCKING"</span>] = <span class="hljs-string">"1"</span><!-- HTML_TAG_END --></pre></div> <h2 class="relative group"><a id="incorrect-output-when-padding-tokens-arent-masked" 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="#incorrect-output-when-padding-tokens-arent-masked"><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>패딩 토큰이 마스킹되지 않은 경우 잘못된 출력(Incorrect output when padding tokens aren’t masked)</span></h2> <p data-svelte-h="svelte-1wwur9b">경우에 따라 <code>input_ids</code>에 패딩 토큰이 포함된 경우 <code>hidden_state</code> 출력이 올바르지 않을 수 있습니다. 데모를 위해 모델과 토크나이저를 가져오세요. 모델의 <code>pad_token_id</code>에 액세스하여 해당 값을 확인할 수 있습니다. 일부 모델의 경우 <code>pad_token_id</code>가 <code>None</code>일 수 있지만 언제든지 수동으로 설정할 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-meta">>>> </span><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> AutoModelForSequenceClassification | |
| <span class="hljs-meta">>>> </span><span class="hljs-keyword">import</span> torch | |
| <span class="hljs-meta">>>> </span>model = AutoModelForSequenceClassification.from_pretrained(<span class="hljs-string">"google-bert/bert-base-uncased"</span>) | |
| <span class="hljs-meta">>>> </span>model.config.pad_token_id | |
| <span class="hljs-number">0</span><!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1qlewzi">다음 예제는 패딩 토큰을 마스킹하지 않은 출력을 보여줍니다:</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-meta">>>> </span>input_ids = torch.tensor([[<span class="hljs-number">7592</span>, <span class="hljs-number">2057</span>, <span class="hljs-number">2097</span>, <span class="hljs-number">2393</span>, <span class="hljs-number">9611</span>, <span class="hljs-number">2115</span>], [<span class="hljs-number">7592</span>, <span class="hljs-number">0</span>, <span class="hljs-number">0</span>, <span class="hljs-number">0</span>, <span class="hljs-number">0</span>, <span class="hljs-number">0</span>]]) | |
| <span class="hljs-meta">>>> </span>output = model(input_ids) | |
| <span class="hljs-meta">>>> </span><span class="hljs-built_in">print</span>(output.logits) | |
| tensor([[ <span class="hljs-number">0.0082</span>, -<span class="hljs-number">0.2307</span>], | |
| [ <span class="hljs-number">0.1317</span>, -<span class="hljs-number">0.1683</span>]], grad_fn=<AddmmBackward0>)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-16yb6is">다음은 두 번째 시퀀스의 실제 출력입니다:</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-meta">>>> </span>input_ids = torch.tensor([[<span class="hljs-number">7592</span>]]) | |
| <span class="hljs-meta">>>> </span>output = model(input_ids) | |
| <span class="hljs-meta">>>> </span><span class="hljs-built_in">print</span>(output.logits) | |
| tensor([[-<span class="hljs-number">0.1008</span>, -<span class="hljs-number">0.4061</span>]], grad_fn=<AddmmBackward0>)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-jqbu9">대부분의 경우 모델에 <code>attention_mask</code>를 제공하여 패딩 토큰을 무시해야 이러한 조용한 오류를 방지할 수 있습니다. 이제 두 번째 시퀀스의 출력이 실제 출력과 일치합니다:</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-1afkfxz">일반적으로 토크나이저는 특정 토크나이저의 기본 값을 기준으로 사용자에 대한 ‘attention_mask’를 만듭니다.</p></div> <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-meta">>>> </span>attention_mask = torch.tensor([[<span class="hljs-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>, <span class="hljs-number">1</span>], [<span class="hljs-number">1</span>, <span class="hljs-number">0</span>, <span class="hljs-number">0</span>, <span class="hljs-number">0</span>, <span class="hljs-number">0</span>, <span class="hljs-number">0</span>]]) | |
| <span class="hljs-meta">>>> </span>output = model(input_ids, attention_mask=attention_mask) | |
| <span class="hljs-meta">>>> </span><span class="hljs-built_in">print</span>(output.logits) | |
| tensor([[ <span class="hljs-number">0.0082</span>, -<span class="hljs-number">0.2307</span>], | |
| [-<span class="hljs-number">0.1008</span>, -<span class="hljs-number">0.4061</span>]], grad_fn=<AddmmBackward0>)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-2020ti">🤗 Transformers는 패딩 토큰이 제공된 경우 패딩 토큰을 마스킹하기 위한 <code>attention_mask</code>를 자동으로 생성하지 않습니다. 그 이유는 다음과 같습니다:</p> <ul data-svelte-h="svelte-iyjkil"><li>일부 모델에는 패딩 토큰이 없습니다.</li> <li>일부 사용 사례의 경우 사용자가 모델이 패딩 토큰을 관리하기를 원합니다.</li></ul> <h2 class="relative group"><a id="valueerror-unrecognized-configuration-class-xyz-for-this-kind-of-automodel" 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="#valueerror-unrecognized-configuration-class-xyz-for-this-kind-of-automodel"><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>ValueError: 이 유형의 AutoModel에 대해 인식할 수 없는 XYZ 구성 클래스(ValueError: Unrecognized configuration class XYZ for this kind of AutoModel)</span></h2> <p data-svelte-h="svelte-psmtvy">일반적으로, 사전 학습된 모델의 인스턴스를 가져오기 위해 <code>AutoModel</code> 클래스를 사용하는 것이 좋습니다. | |
| 이 클래스는 구성에 따라 주어진 체크포인트에서 올바른 아키텍처를 자동으로 추론하고 가져올 수 있습니다. | |
| 모델을 체크포인트에서 가져올 때 이 <code>ValueError</code>가 발생하면, 이는 Auto 클래스가 주어진 체크포인트의 구성에서 | |
| 가져오려는 모델 유형과 매핑을 찾을 수 없다는 것을 의미합니다. 가장 흔하게 발생하는 경우는 | |
| 체크포인트가 주어진 태스크를 지원하지 않을 때입니다. | |
| 예를 들어, 다음 예제에서 질의응답에 대한 GPT2가 없기 때문에 오류가 발생합니다:</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-meta">>>> </span><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> AutoProcessor, AutoModelForQuestionAnswering | |
| <span class="hljs-meta">>>> </span>processor = AutoProcessor.from_pretrained(<span class="hljs-string">"openai-community/gpt2-medium"</span>) | |
| <span class="hljs-meta">>>> </span>model = AutoModelForQuestionAnswering.from_pretrained(<span class="hljs-string">"openai-community/gpt2-medium"</span>) | |
| ValueError: Unrecognized configuration <span class="hljs-keyword">class</span> <<span class="hljs-keyword">class</span> <span class="hljs-string">'transformers.models.gpt2.configuration_gpt2.GPT2Config'</span>> <span class="hljs-keyword">for</span> this kind of AutoModel: AutoModelForQuestionAnswering. | |
| Model <span class="hljs-built_in">type</span> should be one of AlbertConfig, BartConfig, BertConfig, BigBirdConfig, BigBirdPegasusConfig, BloomConfig, ...<!-- 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/transformers/blob/main/docs/source/ko/troubleshooting.md" target="_blank"><span data-svelte-h="svelte-1kd6by1"><</span> <span data-svelte-h="svelte-x0xyl0">></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, 96], | |
| data, | |
| form: null, | |
| error: null | |
| }); | |
| }); | |
| } | |
| </script> | |
Xet Storage Details
- Size:
- 48.3 kB
- Xet hash:
- 8739b52b71f39db058e56f6cc88f1627d73e5c4afb1ba601f616dc6bde4bd2d8
·
Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.