Hugging Face's logo

Buckets:
hf-doc-build
/
doc-dev

Files
xet
hf-doc-build/doc-dev / diffusers /pr_13881 /ko /_app /immutable /nodes /33.6jPkYAaR.js
HuggingFaceDocBuilder's picture
HuggingFaceDocBuilder
download
raw
59.7 kB
import"../chunks/DsnmJJEf.js";import{i as _,h as W,C as O,H as a,a as s,E as S,s as D}from"../chunks/CFM6C53a.js";import{p as Y,o as z,s as l,f as v,a as E,b as x,c as k,n as H}from"../chunks/CNc7KuUZ.js";import{D as q}from"../chunks/BK2xlcGK.js";const K='{"title":"Diffusion 모델을 학습하기","local":"diffusion-모델을-학습하기","sections":[{"title":"학습 구성","local":"학습-구성","sections":[],"depth":2},{"title":"데이터셋 불러오기","local":"데이터셋-불러오기","sections":[],"depth":2},{"title":"UNet2DModel 생성하기","local":"unet2dmodel-생성하기","sections":[],"depth":2},{"title":"스케줄러 생성하기","local":"스케줄러-생성하기","sections":[],"depth":2},{"title":"모델 학습하기","local":"모델-학습하기","sections":[],"depth":2},{"title":"다음 단계","local":"다음-단계","sections":[],"depth":2}],"depth":1}';var L=k('<meta name="hf:doc:metadata"/>'),P=k('<p></p> <!> <!> <!> <p>Unconditional 이미지 생성은 학습에 사용된 데이터셋과 유사한 이미지를 생성하는 diffusion 모델에서 인기 있는 어플리케이션입니다. 일반적으로, 가장 좋은 결과는 특정 데이터셋에 사전 훈련된 모델을 파인튜닝하는 것으로 얻을 수 있습니다. 이 <a href="https://huggingface.co/search/full-text?q=unconditional-image-generation&amp;type=model" rel="nofollow">허브</a>에서 이러한 많은 체크포인트를 찾을 수 있지만, 만약 마음에 드는 체크포인트를 찾지 못했다면, 언제든지 스스로 학습할 수 있습니다!</p> <p>이 튜토리얼은 나만의 🦋 나비 🦋를 생성하기 위해 <a href="https://huggingface.co/datasets/huggan/smithsonian_butterflies_subset" rel="nofollow">Smithsonian Butterflies</a> 데이터셋의 하위 집합에서 <code>UNet2DModel</code> 모델을 학습하는 방법을 가르쳐줄 것입니다.</p> <blockquote class="tip"><p>💡 이 학습 튜토리얼은 <a href="https://colab.research.google.com/github/huggingface/notebooks/blob/main/diffusers/training_example.ipynb" rel="nofollow">Training with 🧨 Diffusers</a> 노트북 기반으로 합니다. Diffusion 모델의 작동 방식 및 자세한 내용은 노트북을 확인하세요!</p></blockquote> <p>시작 전에, 🤗 Datasets을 불러오고 전처리하기 위해 데이터셋이 설치되어 있는지 다수 GPU에서 학습을 간소화하기 위해 🤗 Accelerate 가 설치되어 있는지 확인하세요. 그 후 학습 메트릭을 시각화하기 위해 <a href="https://www.tensorflow.org/tensorboard" rel="nofollow">TensorBoard</a>를 또한 설치하세요. (또한 학습 추적을 위해 <a href="https://docs.wandb.ai/" rel="nofollow">Weights & Biases</a>를 사용할 수 있습니다.)</p> <!> <p>커뮤니티에 모델을 공유할 것을 권장하며, 이를 위해서 Hugging Face 계정에 로그인을 해야 합니다. (계정이 없다면 <a href="https://hf.co/join" rel="nofollow">여기</a>에서 만들 수 있습니다.) 노트북에서 로그인할 수 있으며 메시지가 표시되면 토큰을 입력할 수 있습니다.</p> <!> <p>또는 터미널로 로그인할 수 있습니다:</p> <!> <p>모델 체크포인트가 상당히 크기 때문에 <a href="https://git-lfs.com/" rel="nofollow">Git-LFS</a>에서 대용량 파일의 버전 관리를 할 수 있습니다.</p> <!> <!> <p>편의를 위해 학습 파라미터들을 포함한 <code>TrainingConfig</code> 클래스를 생성합니다 (자유롭게 조정 가능):</p> <!> <!> <p>🤗 Datasets 라이브러리와 <a href="https://huggingface.co/datasets/huggan/smithsonian_butterflies_subset" rel="nofollow">Smithsonian Butterflies</a> 데이터셋을 쉽게 불러올 수 있습니다.</p> <!> <p>💡<a href="https://huggingface.co/huggan" rel="nofollow">HugGan Community Event</a> 에서 추가의 데이터셋을 찾거나 로컬의 <a href="https://huggingface.co/docs/datasets/image_dataset#imagefolder" rel="nofollow"><code>ImageFolder</code></a>를 만듦으로써 나만의 데이터셋을 사용할 수 있습니다. HugGan Community Event 에 가져온 데이터셋의 경우 리포지토리의 id로 <code>config.dataset_name</code> 을 설정하고, 나만의 이미지를 사용하는 경우 <code>imagefolder</code> 를 설정합니다.</p> <p>🤗 Datasets은 <code>Image</code> 기능을 사용해 자동으로 이미지 데이터를 디코딩하고 <a href="https://pillow.readthedocs.io/en/stable/reference/Image.html" rel="nofollow"><code>PIL.Image</code></a>로 불러옵니다. 이를 시각화 해보면:</p> <!> <p><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/butterflies_ds.png"/></p> <p>이미지는 모두 다른 사이즈이기 때문에, 우선 전처리가 필요합니다:</p> <ul><li><code>Resize</code> 는 <code>config.image_size</code> 에 정의된 이미지 사이즈로 변경합니다.</li> <li><code>RandomHorizontalFlip</code> 은 랜덤적으로 이미지를 미러링하여 데이터셋을 보강합니다.</li> <li><code>Normalize</code> 는 모델이 예상하는 [-1, 1] 범위로 픽셀 값을 재조정 하는데 중요합니다.</li></ul> <!> <p>학습 도중에 <code>preprocess</code> 함수를 적용하려면 🤗 Datasets의 <code>set_transform</code> 방법이 사용됩니다.</p> <!> <p>이미지의 크기가 조정되었는지 확인하기 위해 이미지를 다시 시각화해보세요. 이제 <a href="https://pytorch.org/docs/stable/data#torch.utils.data.DataLoader" rel="nofollow">DataLoader</a>에 데이터셋을 포함해 학습할 준비가 되었습니다!</p> <!> <!> <p>🧨 Diffusers에 사전학습된 모델들은 모델 클래스에서 원하는 파라미터로 쉽게 생성할 수 있습니다. 예를 들어, <code>UNet2DModel</code>를 생성하려면:</p> <!> <p>샘플의 이미지 크기와 모델 출력 크기가 맞는지 빠르게 확인하기 위한 좋은 아이디어가 있습니다:</p> <!> <p>훌륭해요! 다음, 이미지에 약간의 노이즈를 더하기 위해 스케줄러가 필요합니다.</p> <!> <p>스케줄러는 모델을 학습 또는 추론에 사용하는지에 따라 다르게 작동합니다. 추론시에, 스케줄러는 노이즈로부터 이미지를 생성합니다. 학습시 스케줄러는 diffusion 과정에서의 특정 포인트로부터 모델의 출력 또는 샘플을 가져와 <em>노이즈 스케줄</em> 과 <em>업데이트 규칙</em>에 따라 이미지에 노이즈를 적용합니다.</p> <p><code>DDPMScheduler</code>를 보면 이전으로부터 <code>sample_image</code>에 랜덤한 노이즈를 더하는 <code>add_noise</code> 메서드를 사용합니다:</p> <!> <p><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/noisy_butterfly.png"/></p> <p>모델의 학습 목적은 이미지에 더해진 노이즈를 예측하는 것입니다. 이 단계에서 손실은 다음과 같이 계산될 수 있습니다:</p> <!> <!> <p>지금까지, 모델 학습을 시작하기 위해 많은 부분을 갖추었으며 이제 남은 것은 모든 것을 조합하는 것입니다.</p> <p>우선 옵티마이저(optimizer)와 학습률 스케줄러(learning rate scheduler)가 필요할 것입니다:</p> <!> <p>그 후, 모델을 평가하는 방법이 필요합니다. 평가를 위해, <code>DDPMPipeline</code>을 사용해 배치의 이미지 샘플들을 생성하고 그리드 형태로 저장할 수 있습니다:</p> <!> <p>TensorBoard에 로깅, 그래디언트 누적 및 혼합 정밀도 학습을 쉽게 수행하기 위해 🤗 Accelerate를 학습 루프에 함께 앞서 말한 모든 구성 정보들을 묶어 진행할 수 있습니다. 허브에 모델을 업로드 하기 위해 리포지토리 이름 및 정보를 가져오기 위한 함수를 작성하고 허브에 업로드할 수 있습니다.</p> <p>💡아래의 학습 루프는 어렵고 길어 보일 수 있지만, 나중에 한 줄의 코드로 학습을 한다면 그만한 가치가 있을 것입니다! 만약 기다리지 못하고 이미지를 생성하고 싶다면, 아래 코드를 자유롭게 붙여넣고 작동시키면 됩니다. 🤗</p> <!> <p>휴, 코드가 꽤 많았네요! 하지만 🤗 Accelerate의 <code>notebook_launcher</code> 함수와 학습을 시작할 준비가 되었습니다. 함수에 학습 루프, 모든 학습 인수, 학습에 사용할 프로세스 수(사용 가능한 GPU의 수를 변경할 수 있음)를 전달합니다:</p> <!> <p>한번 학습이 완료되면, diffusion 모델로 생성된 최종 🦋이미지🦋를 확인해보길 바랍니다!</p> <!> <p><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/butterflies_final.png"/></p> <!> <p>Unconditional 이미지 생성은 학습될 수 있는 작업 중 하나의 예시입니다. 다른 작업과 학습 방법은 <a href="../training/overview">🧨 Diffusers 학습 예시</a> 페이지에서 확인할 수 있습니다. 다음은 학습할 수 있는 몇 가지 예시입니다:</p> <ul><li><a href="../training/text_inversion">Textual Inversion</a>, 특정 시각적 개념을 학습시켜 생성된 이미지에 통합시키는 알고리즘입니다.</li> <li><a href="../training/dreambooth">DreamBooth</a>, 주제에 대한 몇 가지 입력 이미지들이 주어지면 주제에 대한 개인화된 이미지를 생성하기 위한 기술입니다.</li> <li><a href="../training/text2image">Guide</a> 데이터셋에 Stable Diffusion 모델을 파인튜닝하는 방법입니다.</li> <li><a href="../training/lora">Guide</a> LoRA를 사용해 매우 큰 모델을 빠르게 파인튜닝하기 위한 메모리 효율적인 기술입니다.</li></ul> <!> <p></p>',1);function Ml(F,X){Y(X,!1),z(()=>{new URLSearchParams(window.location.search).get("fw")}),_();var M=P();W("hfx23e",G=>{var f=L();D(f,"content",K),E(G,f)});var n=l(v(M),2);O(n,{containerStyle:"float: right; margin-left: 10px; display: inline-flex; position: relative; z-index: 10;"});var U=l(n,2);q(U,{containerStyle:"float: right; margin-left: 10px; display: inline-flex; position: relative; z-index: 10;",options:[{label:"Mixed",value:"https://colab.research.google.com/github/huggingface/notebooks/blob/main/diffusers_doc/ko/basic_training.ipynb"},{label:"PyTorch",value:"https://colab.research.google.com/github/huggingface/notebooks/blob/main/diffusers_doc/ko/pytorch/basic_training.ipynb"},{label:"TensorFlow",value:"https://colab.research.google.com/github/huggingface/notebooks/blob/main/diffusers_doc/ko/tensorflow/basic_training.ipynb"},{label:"Mixed",value:"https://studiolab.sagemaker.aws/import/github/huggingface/notebooks/blob/main/diffusers_doc/ko/basic_training.ipynb"},{label:"PyTorch",value:"https://studiolab.sagemaker.aws/import/github/huggingface/notebooks/blob/main/diffusers_doc/ko/pytorch/basic_training.ipynb"},{label:"TensorFlow",value:"https://studiolab.sagemaker.aws/import/github/huggingface/notebooks/blob/main/diffusers_doc/ko/tensorflow/basic_training.ipynb"}]});var p=l(U,2);a(p,{title:"Diffusion 모델을 학습하기",local:"diffusion-모델을-학습하기",headingTag:"h1"});var e=l(p,10);s(e,{code:"IXBpcCUyMGluc3RhbGwlMjBkaWZmdXNlcnMlNUJ0cmFpbmluZyU1RA==",highlighted:"!pip install diffusers[training]",lang:"bash",wrap:!1});var t=l(e,4);s(t,{code:"ZnJvbSUyMGh1Z2dpbmdmYWNlX2h1YiUyMGltcG9ydCUyMG5vdGVib29rX2xvZ2luJTBBJTBBbm90ZWJvb2tfbG9naW4oKQ==",highlighted:`<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">from</span> huggingface_hub <span class="hljs-keyword">import</span> notebook_login
<span class="hljs-meta">&gt;&gt;&gt; </span>notebook_login()`,lang:"py",wrap:!1});var J=l(t,4);s(J,{code:"aGYlMjBhdXRoJTIwbG9naW4=",highlighted:"hf auth login",lang:"bash",wrap:!1});var c=l(J,4);s(c,{code:"IXN1ZG8lMjBhcHQlMjAtcXElMjBpbnN0YWxsJTIwZ2l0LWxmcyUwQSFnaXQlMjBjb25maWclMjAtLWdsb2JhbCUyMGNyZWRlbnRpYWwuaGVscGVyJTIwc3RvcmU=",highlighted:`!<span class="hljs-built_in">sudo</span> apt -qq install git-lfs
!git config --global credential.helper store`,lang:"bash",wrap:!1});var y=l(c,2);a(y,{title:"학습 구성",local:"학습-구성",headingTag:"h2"});var j=l(y,4);s(j,{code:"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",highlighted:`<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">from</span> dataclasses <span class="hljs-keyword">import</span> dataclass
<span class="hljs-meta">&gt;&gt;&gt; </span>@dataclass
<span class="hljs-meta">... </span><span class="hljs-keyword">class</span> <span class="hljs-title class_">TrainingConfig</span>:
<span class="hljs-meta">... </span> image_size = <span class="hljs-number">128</span> <span class="hljs-comment"># 생성되는 이미지 해상도</span>
<span class="hljs-meta">... </span> train_batch_size = <span class="hljs-number">16</span>
<span class="hljs-meta">... </span> eval_batch_size = <span class="hljs-number">16</span> <span class="hljs-comment"># 평가 동안에 샘플링할 이미지 수</span>
<span class="hljs-meta">... </span> num_epochs = <span class="hljs-number">50</span>
<span class="hljs-meta">... </span> gradient_accumulation_steps = <span class="hljs-number">1</span>
<span class="hljs-meta">... </span> learning_rate = <span class="hljs-number">1e-4</span>
<span class="hljs-meta">... </span> lr_warmup_steps = <span class="hljs-number">500</span>
<span class="hljs-meta">... </span> save_image_epochs = <span class="hljs-number">10</span>
<span class="hljs-meta">... </span> save_model_epochs = <span class="hljs-number">30</span>
<span class="hljs-meta">... </span> mixed_precision = <span class="hljs-string">&quot;fp16&quot;</span> <span class="hljs-comment"># \`no\`는 float32, 자동 혼합 정밀도를 위한 \`fp16\`</span>
<span class="hljs-meta">... </span> output_dir = <span class="hljs-string">&quot;ddpm-butterflies-128&quot;</span> <span class="hljs-comment"># 로컬 및 HF Hub에 저장되는 모델명</span>
<span class="hljs-meta">... </span> push_to_hub = <span class="hljs-literal">True</span> <span class="hljs-comment"># 저장된 모델을 HF Hub에 업로드할지 여부</span>
<span class="hljs-meta">... </span> hub_private_repo = <span class="hljs-literal">None</span>
<span class="hljs-meta">... </span> overwrite_output_dir = <span class="hljs-literal">True</span> <span class="hljs-comment"># 노트북을 다시 실행할 때 이전 모델에 덮어씌울지</span>
<span class="hljs-meta">... </span> seed = <span class="hljs-number">0</span>
<span class="hljs-meta">&gt;&gt;&gt; </span>config = TrainingConfig()`,lang:"py",wrap:!1});var T=l(j,2);a(T,{title:"데이터셋 불러오기",local:"데이터셋-불러오기",headingTag:"h2"});var o=l(T,4);s(o,{code:"ZnJvbSUyMGRhdGFzZXRzJTIwaW1wb3J0JTIwbG9hZF9kYXRhc2V0JTBBJTBBY29uZmlnLmRhdGFzZXRfbmFtZSUyMCUzRCUyMCUyMmh1Z2dhbiUyRnNtaXRoc29uaWFuX2J1dHRlcmZsaWVzX3N1YnNldCUyMiUwQWRhdGFzZXQlMjAlM0QlMjBsb2FkX2RhdGFzZXQoY29uZmlnLmRhdGFzZXRfbmFtZSUyQyUyMHNwbGl0JTNEJTIydHJhaW4lMjIp",highlighted:`<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">from</span> datasets <span class="hljs-keyword">import</span> load_dataset
<span class="hljs-meta">&gt;&gt;&gt; </span>config.dataset_name = <span class="hljs-string">&quot;huggan/smithsonian_butterflies_subset&quot;</span>
<span class="hljs-meta">&gt;&gt;&gt; </span>dataset = load_dataset(config.dataset_name, split=<span class="hljs-string">&quot;train&quot;</span>)`,lang:"py",wrap:!1});var i=l(o,6);s(i,{code:"aW1wb3J0JTIwbWF0cGxvdGxpYi5weXBsb3QlMjBhcyUyMHBsdCUwQSUwQWZpZyUyQyUyMGF4cyUyMCUzRCUyMHBsdC5zdWJwbG90cygxJTJDJTIwNCUyQyUyMGZpZ3NpemUlM0QoMTYlMkMlMjA0KSklMEFmb3IlMjBpJTJDJTIwaW1hZ2UlMjBpbiUyMGVudW1lcmF0ZShkYXRhc2V0JTVCJTNBNCU1RCU1QiUyMmltYWdlJTIyJTVEKSUzQSUwQSUyMCUyMCUyMCUyMGF4cyU1QmklNUQuaW1zaG93KGltYWdlKSUwQSUyMCUyMCUyMCUyMGF4cyU1QmklNUQuc2V0X2F4aXNfb2ZmKCklMEFmaWcuc2hvdygp",highlighted:`<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">import</span> matplotlib.pyplot <span class="hljs-keyword">as</span> plt
<span class="hljs-meta">&gt;&gt;&gt; </span>fig, axs = plt.subplots(<span class="hljs-number">1</span>, <span class="hljs-number">4</span>, figsize=(<span class="hljs-number">16</span>, <span class="hljs-number">4</span>))
<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">for</span> i, image <span class="hljs-keyword">in</span> <span class="hljs-built_in">enumerate</span>(dataset[:<span class="hljs-number">4</span>][<span class="hljs-string">&quot;image&quot;</span>]):
<span class="hljs-meta">... </span> axs[i].imshow(image)
<span class="hljs-meta">... </span> axs[i].set_axis_off()
<span class="hljs-meta">&gt;&gt;&gt; </span>fig.show()`,lang:"py",wrap:!1});var h=l(i,8);s(h,{code:"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",highlighted:`<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">from</span> torchvision <span class="hljs-keyword">import</span> transforms
<span class="hljs-meta">&gt;&gt;&gt; </span>preprocess = transforms.Compose(
<span class="hljs-meta">... </span> [
<span class="hljs-meta">... </span> transforms.Resize((config.image_size, config.image_size)),
<span class="hljs-meta">... </span> transforms.RandomHorizontalFlip(),
<span class="hljs-meta">... </span> transforms.ToTensor(),
<span class="hljs-meta">... </span> transforms.Normalize([<span class="hljs-number">0.5</span>], [<span class="hljs-number">0.5</span>]),
<span class="hljs-meta">... </span> ]
<span class="hljs-meta">... </span>)`,lang:"py",wrap:!1});var w=l(h,4);s(w,{code:"ZGVmJTIwdHJhbnNmb3JtKGV4YW1wbGVzKSUzQSUwQSUyMCUyMCUyMCUyMGltYWdlcyUyMCUzRCUyMCU1QnByZXByb2Nlc3MoaW1hZ2UuY29udmVydCglMjJSR0IlMjIpKSUyMGZvciUyMGltYWdlJTIwaW4lMjBleGFtcGxlcyU1QiUyMmltYWdlJTIyJTVEJTVEJTBBJTIwJTIwJTIwJTIwcmV0dXJuJTIwJTdCJTIyaW1hZ2VzJTIyJTNBJTIwaW1hZ2VzJTdEJTBBJTBBJTBBZGF0YXNldC5zZXRfdHJhbnNmb3JtKHRyYW5zZm9ybSk=",highlighted:`<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">def</span> <span class="hljs-title function_">transform</span>(<span class="hljs-params">examples</span>):
<span class="hljs-meta">... </span> images = [preprocess(image.convert(<span class="hljs-string">&quot;RGB&quot;</span>)) <span class="hljs-keyword">for</span> image <span class="hljs-keyword">in</span> examples[<span class="hljs-string">&quot;image&quot;</span>]]
<span class="hljs-meta">... </span> <span class="hljs-keyword">return</span> {<span class="hljs-string">&quot;images&quot;</span>: images}
<span class="hljs-meta">&gt;&gt;&gt; </span>dataset.set_transform(transform)`,lang:"py",wrap:!1});var C=l(w,4);s(C,{code:"aW1wb3J0JTIwdG9yY2glMEElMEF0cmFpbl9kYXRhbG9hZGVyJTIwJTNEJTIwdG9yY2gudXRpbHMuZGF0YS5EYXRhTG9hZGVyKGRhdGFzZXQlMkMlMjBiYXRjaF9zaXplJTNEY29uZmlnLnRyYWluX2JhdGNoX3NpemUlMkMlMjBzaHVmZmxlJTNEVHJ1ZSk=",highlighted:`<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">import</span> torch
<span class="hljs-meta">&gt;&gt;&gt; </span>train_dataloader = torch.utils.data.DataLoader(dataset, batch_size=config.train_batch_size, shuffle=<span class="hljs-literal">True</span>)`,lang:"py",wrap:!1});var m=l(C,2);a(m,{title:"UNet2DModel 생성하기",local:"unet2dmodel-생성하기",headingTag:"h2"});var I=l(m,4);s(I,{code:"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",highlighted:`<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> UNet2DModel
<span class="hljs-meta">&gt;&gt;&gt; </span>model = UNet2DModel(
<span class="hljs-meta">... </span> sample_size=config.image_size, <span class="hljs-comment"># 타겟 이미지 해상도</span>
<span class="hljs-meta">... </span> in_channels=<span class="hljs-number">3</span>, <span class="hljs-comment"># 입력 채널 수, RGB 이미지에서 3</span>
<span class="hljs-meta">... </span> out_channels=<span class="hljs-number">3</span>, <span class="hljs-comment"># 출력 채널 수</span>
<span class="hljs-meta">... </span> layers_per_block=<span class="hljs-number">2</span>, <span class="hljs-comment"># UNet 블럭당 몇 개의 ResNet 레이어가 사용되는지</span>
<span class="hljs-meta">... </span> block_out_channels=(<span class="hljs-number">128</span>, <span class="hljs-number">128</span>, <span class="hljs-number">256</span>, <span class="hljs-number">256</span>, <span class="hljs-number">512</span>, <span class="hljs-number">512</span>), <span class="hljs-comment"># 각 UNet 블럭을 위한 출력 채널 수</span>
<span class="hljs-meta">... </span> down_block_types=(
<span class="hljs-meta">... </span> <span class="hljs-string">&quot;DownBlock2D&quot;</span>, <span class="hljs-comment"># 일반적인 ResNet 다운샘플링 블럭</span>
<span class="hljs-meta">... </span> <span class="hljs-string">&quot;DownBlock2D&quot;</span>,
<span class="hljs-meta">... </span> <span class="hljs-string">&quot;DownBlock2D&quot;</span>,
<span class="hljs-meta">... </span> <span class="hljs-string">&quot;DownBlock2D&quot;</span>,
<span class="hljs-meta">... </span> <span class="hljs-string">&quot;AttnDownBlock2D&quot;</span>, <span class="hljs-comment"># spatial self-attention이 포함된 일반적인 ResNet 다운샘플링 블럭</span>
<span class="hljs-meta">... </span> <span class="hljs-string">&quot;DownBlock2D&quot;</span>,
<span class="hljs-meta">... </span> ),
<span class="hljs-meta">... </span> up_block_types=(
<span class="hljs-meta">... </span> <span class="hljs-string">&quot;UpBlock2D&quot;</span>, <span class="hljs-comment"># 일반적인 ResNet 업샘플링 블럭</span>
<span class="hljs-meta">... </span> <span class="hljs-string">&quot;AttnUpBlock2D&quot;</span>, <span class="hljs-comment"># spatial self-attention이 포함된 일반적인 ResNet 업샘플링 블럭</span>
<span class="hljs-meta">... </span> <span class="hljs-string">&quot;UpBlock2D&quot;</span>,
<span class="hljs-meta">... </span> <span class="hljs-string">&quot;UpBlock2D&quot;</span>,
<span class="hljs-meta">... </span> <span class="hljs-string">&quot;UpBlock2D&quot;</span>,
<span class="hljs-meta">... </span> <span class="hljs-string">&quot;UpBlock2D&quot;</span>,
<span class="hljs-meta">... </span> ),
<span class="hljs-meta">... </span>)`,lang:"py",wrap:!1});var r=l(I,4);s(r,{code:"c2FtcGxlX2ltYWdlJTIwJTNEJTIwZGF0YXNldCU1QjAlNUQlNUIlMjJpbWFnZXMlMjIlNUQudW5zcXVlZXplKDApJTBBcHJpbnQoJTIySW5wdXQlMjBzaGFwZSUzQSUyMiUyQyUyMHNhbXBsZV9pbWFnZS5zaGFwZSklMEElMEFwcmludCglMjJPdXRwdXQlMjBzaGFwZSUzQSUyMiUyQyUyMG1vZGVsKHNhbXBsZV9pbWFnZSUyQyUyMHRpbWVzdGVwJTNEMCkuc2FtcGxlLnNoYXBlKQ==",highlighted:`<span class="hljs-meta">&gt;&gt;&gt; </span>sample_image = dataset[<span class="hljs-number">0</span>][<span class="hljs-string">&quot;images&quot;</span>].unsqueeze(<span class="hljs-number">0</span>)
<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-built_in">print</span>(<span class="hljs-string">&quot;Input shape:&quot;</span>, sample_image.shape)
Input shape: torch.Size([<span class="hljs-number">1</span>, <span class="hljs-number">3</span>, <span class="hljs-number">128</span>, <span class="hljs-number">128</span>])
<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-built_in">print</span>(<span class="hljs-string">&quot;Output shape:&quot;</span>, model(sample_image, timestep=<span class="hljs-number">0</span>).sample.shape)
Output shape: torch.Size([<span class="hljs-number">1</span>, <span class="hljs-number">3</span>, <span class="hljs-number">128</span>, <span class="hljs-number">128</span>])`,lang:"py",wrap:!1});var g=l(r,4);a(g,{title:"스케줄러 생성하기",local:"스케줄러-생성하기",headingTag:"h2"});var d=l(g,6);s(d,{code:"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",highlighted:`<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">import</span> torch
<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">from</span> PIL <span class="hljs-keyword">import</span> Image
<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> DDPMScheduler
<span class="hljs-meta">&gt;&gt;&gt; </span>noise_scheduler = DDPMScheduler(num_train_timesteps=<span class="hljs-number">1000</span>)
<span class="hljs-meta">&gt;&gt;&gt; </span>noise = torch.randn(sample_image.shape)
<span class="hljs-meta">&gt;&gt;&gt; </span>timesteps = torch.LongTensor([<span class="hljs-number">50</span>])
<span class="hljs-meta">&gt;&gt;&gt; </span>noisy_image = noise_scheduler.add_noise(sample_image, noise, timesteps)
<span class="hljs-meta">&gt;&gt;&gt; </span>Image.fromarray(((noisy_image.permute(<span class="hljs-number">0</span>, <span class="hljs-number">2</span>, <span class="hljs-number">3</span>, <span class="hljs-number">1</span>) + <span class="hljs-number">1.0</span>) * <span class="hljs-number">127.5</span>).<span class="hljs-built_in">type</span>(torch.uint8).numpy()[<span class="hljs-number">0</span>])`,lang:"py",wrap:!1});var Q=l(d,6);s(Q,{code:"aW1wb3J0JTIwdG9yY2gubm4uZnVuY3Rpb25hbCUyMGFzJTIwRiUwQSUwQW5vaXNlX3ByZWQlMjAlM0QlMjBtb2RlbChub2lzeV9pbWFnZSUyQyUyMHRpbWVzdGVwcykuc2FtcGxlJTBBbG9zcyUyMCUzRCUyMEYubXNlX2xvc3Mobm9pc2VfcHJlZCUyQyUyMG5vaXNlKQ==",highlighted:`<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">import</span> torch.nn.functional <span class="hljs-keyword">as</span> F
<span class="hljs-meta">&gt;&gt;&gt; </span>noise_pred = model(noisy_image, timesteps).sample
<span class="hljs-meta">&gt;&gt;&gt; </span>loss = F.mse_loss(noise_pred, noise)`,lang:"py",wrap:!1});var V=l(Q,2);a(V,{title:"모델 학습하기",local:"모델-학습하기",headingTag:"h2"});var b=l(V,6);s(b,{code:"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",highlighted:`<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">from</span> diffusers.optimization <span class="hljs-keyword">import</span> get_cosine_schedule_with_warmup
<span class="hljs-meta">&gt;&gt;&gt; </span>optimizer = torch.optim.AdamW(model.parameters(), lr=config.learning_rate)
<span class="hljs-meta">&gt;&gt;&gt; </span>lr_scheduler = get_cosine_schedule_with_warmup(
<span class="hljs-meta">... </span> optimizer=optimizer,
<span class="hljs-meta">... </span> num_warmup_steps=config.lr_warmup_steps,
<span class="hljs-meta">... </span> num_training_steps=(<span class="hljs-built_in">len</span>(train_dataloader) * config.num_epochs),
<span class="hljs-meta">... </span>)`,lang:"py",wrap:!1});var u=l(b,4);s(u,{code:"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",highlighted:`<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> DDPMPipeline
<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">import</span> math
<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">import</span> os
<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">def</span> <span class="hljs-title function_">make_grid</span>(<span class="hljs-params">images, rows, cols</span>):
<span class="hljs-meta">... </span> w, h = images[<span class="hljs-number">0</span>].size
<span class="hljs-meta">... </span> grid = Image.new(<span class="hljs-string">&quot;RGB&quot;</span>, size=(cols * w, rows * h))
<span class="hljs-meta">... </span> <span class="hljs-keyword">for</span> i, image <span class="hljs-keyword">in</span> <span class="hljs-built_in">enumerate</span>(images):
<span class="hljs-meta">... </span> grid.paste(image, box=(i % cols * w, i // cols * h))
<span class="hljs-meta">... </span> <span class="hljs-keyword">return</span> grid
<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">def</span> <span class="hljs-title function_">evaluate</span>(<span class="hljs-params">config, epoch, pipeline</span>):
<span class="hljs-meta">... </span> <span class="hljs-comment"># 랜덤한 노이즈로 부터 이미지를 추출합니다.(이는 역전파 diffusion 과정입니다.)</span>
<span class="hljs-meta">... </span> <span class="hljs-comment"># 기본 파이프라인 출력 형태는 \`List[PIL.Image]\` 입니다.</span>
<span class="hljs-meta">... </span> images = pipeline(
<span class="hljs-meta">... </span> batch_size=config.eval_batch_size,
<span class="hljs-meta">... </span> generator=torch.manual_seed(config.seed),
<span class="hljs-meta">... </span> ).images
<span class="hljs-meta">... </span> <span class="hljs-comment"># 이미지들을 그리드로 만들어줍니다.</span>
<span class="hljs-meta">... </span> image_grid = make_grid(images, rows=<span class="hljs-number">4</span>, cols=<span class="hljs-number">4</span>)
<span class="hljs-meta">... </span> <span class="hljs-comment"># 이미지들을 저장합니다.</span>
<span class="hljs-meta">... </span> test_dir = os.path.join(config.output_dir, <span class="hljs-string">&quot;samples&quot;</span>)
<span class="hljs-meta">... </span> os.makedirs(test_dir, exist_ok=<span class="hljs-literal">True</span>)
<span class="hljs-meta">... </span> image_grid.save(<span class="hljs-string">f&quot;<span class="hljs-subst">{test_dir}</span>/<span class="hljs-subst">{epoch:04d}</span>.png&quot;</span>)`,lang:"py",wrap:!1});var A=l(u,6);s(A,{code:"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",highlighted:`<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">from</span> accelerate <span class="hljs-keyword">import</span> Accelerator
<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">from</span> huggingface_hub <span class="hljs-keyword">import</span> create_repo, upload_folder
<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">from</span> tqdm.auto <span class="hljs-keyword">import</span> tqdm
<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">from</span> pathlib <span class="hljs-keyword">import</span> Path
<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">import</span> os
<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">def</span> <span class="hljs-title function_">train_loop</span>(<span class="hljs-params">config, model, noise_scheduler, optimizer, train_dataloader, lr_scheduler</span>):
<span class="hljs-meta">... </span> <span class="hljs-comment"># Initialize accelerator and tensorboard logging</span>
<span class="hljs-meta">... </span> accelerator = Accelerator(
<span class="hljs-meta">... </span> mixed_precision=config.mixed_precision,
<span class="hljs-meta">... </span> gradient_accumulation_steps=config.gradient_accumulation_steps,
<span class="hljs-meta">... </span> log_with=<span class="hljs-string">&quot;tensorboard&quot;</span>,
<span class="hljs-meta">... </span> project_dir=os.path.join(config.output_dir, <span class="hljs-string">&quot;logs&quot;</span>),
<span class="hljs-meta">... </span> )
<span class="hljs-meta">... </span> <span class="hljs-keyword">if</span> accelerator.is_main_process:
<span class="hljs-meta">... </span> <span class="hljs-keyword">if</span> config.output_dir <span class="hljs-keyword">is</span> <span class="hljs-keyword">not</span> <span class="hljs-literal">None</span>:
<span class="hljs-meta">... </span> os.makedirs(config.output_dir, exist_ok=<span class="hljs-literal">True</span>)
<span class="hljs-meta">... </span> <span class="hljs-keyword">if</span> config.push_to_hub:
<span class="hljs-meta">... </span> repo_id = create_repo(
<span class="hljs-meta">... </span> repo_id=config.hub_model_id <span class="hljs-keyword">or</span> Path(config.output_dir).name, exist_ok=<span class="hljs-literal">True</span>
<span class="hljs-meta">... </span> ).repo_id
<span class="hljs-meta">... </span> accelerator.init_trackers(<span class="hljs-string">&quot;train_example&quot;</span>)
<span class="hljs-meta">... </span> <span class="hljs-comment"># 모든 것이 준비되었습니다.</span>
<span class="hljs-meta">... </span> <span class="hljs-comment"># 기억해야 할 특정한 순서는 없으며 준비한 방법에 제공한 것과 동일한 순서로 객체의 압축을 풀면 됩니다.</span>
<span class="hljs-meta">... </span> model, optimizer, train_dataloader, lr_scheduler = accelerator.prepare(
<span class="hljs-meta">... </span> model, optimizer, train_dataloader, lr_scheduler
<span class="hljs-meta">... </span> )
<span class="hljs-meta">... </span> global_step = <span class="hljs-number">0</span>
<span class="hljs-meta">... </span> <span class="hljs-comment"># 이제 모델을 학습합니다.</span>
<span class="hljs-meta">... </span> <span class="hljs-keyword">for</span> epoch <span class="hljs-keyword">in</span> <span class="hljs-built_in">range</span>(config.num_epochs):
<span class="hljs-meta">... </span> progress_bar = tqdm(total=<span class="hljs-built_in">len</span>(train_dataloader), disable=<span class="hljs-keyword">not</span> accelerator.is_local_main_process)
<span class="hljs-meta">... </span> progress_bar.set_description(<span class="hljs-string">f&quot;Epoch <span class="hljs-subst">{epoch}</span>&quot;</span>)
<span class="hljs-meta">... </span> <span class="hljs-keyword">for</span> step, batch <span class="hljs-keyword">in</span> <span class="hljs-built_in">enumerate</span>(train_dataloader):
<span class="hljs-meta">... </span> clean_images = batch[<span class="hljs-string">&quot;images&quot;</span>]
<span class="hljs-meta">... </span> <span class="hljs-comment"># 이미지에 더할 노이즈를 샘플링합니다.</span>
<span class="hljs-meta">... </span> noise = torch.randn(clean_images.shape, device=clean_images.device)
<span class="hljs-meta">... </span> bs = clean_images.shape[<span class="hljs-number">0</span>]
<span class="hljs-meta">... </span> <span class="hljs-comment"># 각 이미지를 위한 랜덤한 타임스텝(timestep)을 샘플링합니다.</span>
<span class="hljs-meta">... </span> timesteps = torch.randint(
<span class="hljs-meta">... </span> <span class="hljs-number">0</span>, noise_scheduler.config.num_train_timesteps, (bs,), device=clean_images.device,
<span class="hljs-meta">... </span> dtype=torch.int64
<span class="hljs-meta">... </span> )
<span class="hljs-meta">... </span> <span class="hljs-comment"># 각 타임스텝의 노이즈 크기에 따라 깨끗한 이미지에 노이즈를 추가합니다.</span>
<span class="hljs-meta">... </span> <span class="hljs-comment"># (이는 foward diffusion 과정입니다.)</span>
<span class="hljs-meta">... </span> noisy_images = noise_scheduler.add_noise(clean_images, noise, timesteps)
<span class="hljs-meta">... </span> <span class="hljs-keyword">with</span> accelerator.accumulate(model):
<span class="hljs-meta">... </span> <span class="hljs-comment"># 노이즈를 반복적으로 예측합니다.</span>
<span class="hljs-meta">... </span> noise_pred = model(noisy_images, timesteps, return_dict=<span class="hljs-literal">False</span>)[<span class="hljs-number">0</span>]
<span class="hljs-meta">... </span> loss = F.mse_loss(noise_pred, noise)
<span class="hljs-meta">... </span> accelerator.backward(loss)
<span class="hljs-meta">... </span> accelerator.clip_grad_norm_(model.parameters(), <span class="hljs-number">1.0</span>)
<span class="hljs-meta">... </span> optimizer.step()
<span class="hljs-meta">... </span> lr_scheduler.step()
<span class="hljs-meta">... </span> optimizer.zero_grad()
<span class="hljs-meta">... </span> progress_bar.update(<span class="hljs-number">1</span>)
<span class="hljs-meta">... </span> logs = {<span class="hljs-string">&quot;loss&quot;</span>: loss.detach().item(), <span class="hljs-string">&quot;lr&quot;</span>: lr_scheduler.get_last_lr()[<span class="hljs-number">0</span>], <span class="hljs-string">&quot;step&quot;</span>: global_step}
<span class="hljs-meta">... </span> progress_bar.set_postfix(**logs)
<span class="hljs-meta">... </span> accelerator.log(logs, step=global_step)
<span class="hljs-meta">... </span> global_step += <span class="hljs-number">1</span>
<span class="hljs-meta">... </span> <span class="hljs-comment"># 각 에포크가 끝난 후 evaluate()와 몇 가지 데모 이미지를 선택적으로 샘플링하고 모델을 저장합니다.</span>
<span class="hljs-meta">... </span> <span class="hljs-keyword">if</span> accelerator.is_main_process:
<span class="hljs-meta">... </span> pipeline = DDPMPipeline(unet=accelerator.unwrap_model(model), scheduler=noise_scheduler)
<span class="hljs-meta">... </span> <span class="hljs-keyword">if</span> (epoch + <span class="hljs-number">1</span>) % config.save_image_epochs == <span class="hljs-number">0</span> <span class="hljs-keyword">or</span> epoch == config.num_epochs - <span class="hljs-number">1</span>:
<span class="hljs-meta">... </span> evaluate(config, epoch, pipeline)
<span class="hljs-meta">... </span> <span class="hljs-keyword">if</span> (epoch + <span class="hljs-number">1</span>) % config.save_model_epochs == <span class="hljs-number">0</span> <span class="hljs-keyword">or</span> epoch == config.num_epochs - <span class="hljs-number">1</span>:
<span class="hljs-meta">... </span> <span class="hljs-keyword">if</span> config.push_to_hub:
<span class="hljs-meta">... </span> upload_folder(
<span class="hljs-meta">... </span> repo_id=repo_id,
<span class="hljs-meta">... </span> folder_path=config.output_dir,
<span class="hljs-meta">... </span> commit_message=<span class="hljs-string">f&quot;Epoch <span class="hljs-subst">{epoch}</span>&quot;</span>,
<span class="hljs-meta">... </span> ignore_patterns=[<span class="hljs-string">&quot;step_*&quot;</span>, <span class="hljs-string">&quot;epoch_*&quot;</span>],
<span class="hljs-meta">... </span> )
<span class="hljs-meta">... </span> <span class="hljs-keyword">else</span>:
<span class="hljs-meta">... </span> pipeline.save_pretrained(config.output_dir)`,lang:"py",wrap:!1});var R=l(A,4);s(R,{code:"ZnJvbSUyMGFjY2VsZXJhdGUlMjBpbXBvcnQlMjBub3RlYm9va19sYXVuY2hlciUwQSUwQWFyZ3MlMjAlM0QlMjAoY29uZmlnJTJDJTIwbW9kZWwlMkMlMjBub2lzZV9zY2hlZHVsZXIlMkMlMjBvcHRpbWl6ZXIlMkMlMjB0cmFpbl9kYXRhbG9hZGVyJTJDJTIwbHJfc2NoZWR1bGVyKSUwQSUwQW5vdGVib29rX2xhdW5jaGVyKHRyYWluX2xvb3AlMkMlMjBhcmdzJTJDJTIwbnVtX3Byb2Nlc3NlcyUzRDEp",highlighted:`<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">from</span> accelerate <span class="hljs-keyword">import</span> notebook_launcher
<span class="hljs-meta">&gt;&gt;&gt; </span>args = (config, model, noise_scheduler, optimizer, train_dataloader, lr_scheduler)
<span class="hljs-meta">&gt;&gt;&gt; </span>notebook_launcher(train_loop, args, num_processes=<span class="hljs-number">1</span>)`,lang:"py",wrap:!1});var B=l(R,4);s(B,{code:"aW1wb3J0JTIwZ2xvYiUwQSUwQXNhbXBsZV9pbWFnZXMlMjAlM0QlMjBzb3J0ZWQoZ2xvYi5nbG9iKGYlMjIlN0Jjb25maWcub3V0cHV0X2RpciU3RCUyRnNhbXBsZXMlMkYqLnBuZyUyMikpJTBBSW1hZ2Uub3BlbihzYW1wbGVfaW1hZ2VzJTVCLTElNUQp",highlighted:`<span class="hljs-meta">&gt;&gt;&gt; </span><span class="hljs-keyword">import</span> glob
<span class="hljs-meta">&gt;&gt;&gt; </span>sample_images = <span class="hljs-built_in">sorted</span>(glob.glob(<span class="hljs-string">f&quot;<span class="hljs-subst">{config.output_dir}</span>/samples/*.png&quot;</span>))
<span class="hljs-meta">&gt;&gt;&gt; </span>Image.<span class="hljs-built_in">open</span>(sample_images[-<span class="hljs-number">1</span>])`,lang:"py",wrap:!1});var Z=l(B,4);a(Z,{title:"다음 단계",local:"다음-단계",headingTag:"h2"});var N=l(Z,6);S(N,{source:"https://github.com/huggingface/diffusers/blob/main/docs/source/ko/tutorials/basic_training.md"}),H(2),E(F,M),x()}export{Ml as component};

Xet Storage Details

Size:
59.7 kB
·
Xet hash:
dbb838ae40eb6a3410c569257e1a04bc23d5eecb4bb3ea66ee532a17b9b63072

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