docs/beginner/vt_tutorial.html



<!DOCTYPE html>
<!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
<!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
<head>
  <meta charset="utf-8">
  
  <meta name="viewport" content="width=device-width, initial-scale=1.0">
  
  <title>배포를 위한 비전 트랜스포머(Vision Transformer) 모델 최적화하기 &mdash; PyTorch Tutorials 1.10.2+cu102 documentation</title>
  

  <link rel="stylesheet" href="../_static/css/theme.css" type="text/css" />
  <!-- <link rel="stylesheet" href="../_static/pygments.css" type="text/css" /> -->
  <link rel="stylesheet" href="../_static/copybutton.css" type="text/css" />
  <link rel="stylesheet" href="../_static/gallery.css" type="text/css" />
  <link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/katex@0.10.0-beta/dist/katex.min.css" type="text/css" />
  <link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/katex@0.13.11/dist/katex.min.css" type="text/css" />
  <link rel="stylesheet" href="../_static/katex-math.css" type="text/css" />
    <link rel="index" title="Index" href="../genindex.html" />
    <link rel="search" title="Search" href="../search.html" />
    <link rel="next" title="Audio I/O" href="audio_io_tutorial.html" />
    <link rel="prev" title="DCGAN 튜토리얼" href="dcgan_faces_tutorial.html" /> 

  
  <script src="../_static/js/modernizr.min.js"></script>

  <!-- Preload the theme fonts -->

<link rel="preload" href="../_static/fonts/FreightSans/freight-sans-book.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="../_static/fonts/FreightSans/freight-sans-medium.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="../_static/fonts/IBMPlexMono/IBMPlexMono-Medium.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="../_static/fonts/FreightSans/freight-sans-bold.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="../_static/fonts/FreightSans/freight-sans-medium-italic.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="../_static/fonts/IBMPlexMono/IBMPlexMono-SemiBold.woff2" as="font" type="font/woff2" crossorigin="anonymous">

<!-- Preload the katex fonts -->

<link rel="preload" href="https://cdn.jsdelivr.net/npm/katex@0.10.0/dist/fonts/KaTeX_Math-Italic.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/katex@0.10.0/dist/fonts/KaTeX_Main-Regular.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/katex@0.10.0/dist/fonts/KaTeX_Main-Bold.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/katex@0.10.0/dist/fonts/KaTeX_Size1-Regular.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/katex@0.10.0/dist/fonts/KaTeX_Size4-Regular.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/katex@0.10.0/dist/fonts/KaTeX_Size2-Regular.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/katex@0.10.0/dist/fonts/KaTeX_Size3-Regular.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/katex@0.10.0/dist/fonts/KaTeX_Caligraphic-Regular.woff2" as="font" type="font/woff2" crossorigin="anonymous">
  <link rel="stylesheet" href="https://use.fontawesome.com/releases/v5.15.2/css/all.css" integrity="sha384-vSIIfh2YWi9wW0r9iZe7RJPrKwp6bG+s9QZMoITbCckVJqGCCRhc+ccxNcdpHuYu" crossorigin="anonymous">
</head>

<div class="container-fluid header-holder tutorials-header" id="header-holder">
  <div class="container">
    <div class="header-container">
      <a class="header-logo" href="https://pytorch.kr/" aria-label="PyTorch"></a>

      <div class="main-menu">
        <ul>
          <li>
            <a href="https://pytorch.kr/get-started">시작하기</a>
          </li>

          <li class="active">
            <a href="https://tutorials.pytorch.kr">튜토리얼</a>
          </li>

          <li>
            <a href="https://pytorch.kr/hub">허브</a>
          </li>

          <li>
            <a href="https://discuss.pytorch.kr">커뮤니티</a>
          </li>
        </ul>
      </div>

      <a class="main-menu-open-button" href="#" data-behavior="open-mobile-menu"></a>
    </div>
  </div>
</div>

<body class="pytorch-body">

   
    <div class="table-of-contents-link-wrapper">
      <span>Table of Contents</span>
      <a href="#" class="toggle-table-of-contents" data-behavior="toggle-table-of-contents"></a>
    </div>

    <nav data-toggle="wy-nav-shift" class="pytorch-left-menu" id="pytorch-left-menu">
      <div class="pytorch-side-scroll">
        <div class="pytorch-menu pytorch-menu-vertical" data-spy="affix" role="navigation" aria-label="main navigation">
          <div class="pytorch-left-menu-search">
            

                <div class="version">
                  1.10.2+cu102
                </div>
              
            
<div role="search">
  <form id="rtd-search-form" class="wy-form" action="../search.html" method="get">
    <input type="text" name="q" placeholder="Search Tutorials" />
    <input type="hidden" name="check_keywords" value="yes" />
    <input type="hidden" name="area" value="default" />
  </form>
</div>

            
          </div>

          
              <p class="caption"><span class="caption-text">파이토치(PyTorch) 레시피</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../recipes/recipes_index.html">모든 레시피 보기</a></li>
<li class="toctree-l1"><a class="reference internal" href="../prototype/prototype_index.html">모든 프로토타입 레시피 보기</a></li>
</ul>
<p class="caption"><span class="caption-text">파이토치(PyTorch) 시작하기</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="basics/intro.html">파이토치(PyTorch) 기본 익히기</a></li>
<li class="toctree-l1"><a class="reference internal" href="basics/quickstart_tutorial.html">빠른 시작(Quickstart)</a></li>
<li class="toctree-l1"><a class="reference internal" href="basics/tensorqs_tutorial.html">텐서(Tensor)</a></li>
<li class="toctree-l1"><a class="reference internal" href="basics/data_tutorial.html">Dataset과 DataLoader</a></li>
<li class="toctree-l1"><a class="reference internal" href="basics/transforms_tutorial.html">변형(Transform)</a></li>
<li class="toctree-l1"><a class="reference internal" href="basics/buildmodel_tutorial.html">신경망 모델 구성하기</a></li>
<li class="toctree-l1"><a class="reference internal" href="basics/autogradqs_tutorial.html"><code class="docutils literal notranslate"><span class="pre">torch.autograd</span></code>를 사용한 자동 미분</a></li>
<li class="toctree-l1"><a class="reference internal" href="basics/optimization_tutorial.html">모델 매개변수 최적화하기</a></li>
<li class="toctree-l1"><a class="reference internal" href="basics/saveloadrun_tutorial.html">모델 저장하고 불러오기</a></li>
</ul>
<p class="caption"><span class="caption-text">Introduction to PyTorch on YouTube</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="introyt.html">Introduction to PyTorch - YouTube Series</a></li>
<li class="toctree-l1"><a class="reference internal" href="introyt/introyt1_tutorial.html">Introduction to PyTorch</a></li>
<li class="toctree-l1"><a class="reference internal" href="introyt/tensors_deeper_tutorial.html">Introduction to PyTorch Tensors</a></li>
<li class="toctree-l1"><a class="reference internal" href="introyt/autogradyt_tutorial.html">The Fundamentals of Autograd</a></li>
<li class="toctree-l1"><a class="reference internal" href="introyt/modelsyt_tutorial.html">Building Models with PyTorch</a></li>
<li class="toctree-l1"><a class="reference internal" href="introyt/tensorboardyt_tutorial.html">PyTorch TensorBoard Support</a></li>
<li class="toctree-l1"><a class="reference internal" href="introyt/trainingyt.html">Training with PyTorch</a></li>
<li class="toctree-l1"><a class="reference internal" href="introyt/captumyt.html">Model Understanding with Captum</a></li>
</ul>
<p class="caption"><span class="caption-text">파이토치(PyTorch) 배우기</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="deep_learning_60min_blitz.html">PyTorch로 딥러닝하기: 60분만에 끝장내기</a></li>
<li class="toctree-l1"><a class="reference internal" href="pytorch_with_examples.html">예제로 배우는 파이토치(PyTorch)</a></li>
<li class="toctree-l1"><a class="reference internal" href="nn_tutorial.html"><cite>torch.nn</cite> 이 <em>실제로</em> 무엇인가요?</a></li>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/tensorboard_tutorial.html">TensorBoard로 모델, 데이터, 학습 시각화하기</a></li>
</ul>
<p class="caption"><span class="caption-text">이미지/비디오</span></p>
<ul class="current">
<li class="toctree-l1"><a class="reference internal" href="../intermediate/torchvision_tutorial.html">TorchVision 객체 검출 미세조정(Finetuning) 튜토리얼</a></li>
<li class="toctree-l1"><a class="reference internal" href="transfer_learning_tutorial.html">컴퓨터 비전(Vision)을 위한 전이학습(Transfer Learning)</a></li>
<li class="toctree-l1"><a class="reference internal" href="fgsm_tutorial.html">적대적 예제 생성(Adversarial Example Generation)</a></li>
<li class="toctree-l1"><a class="reference internal" href="dcgan_faces_tutorial.html">DCGAN 튜토리얼</a></li>
<li class="toctree-l1 current"><a class="current reference internal" href="#">배포를 위한 비전 트랜스포머(Vision Transformer) 모델 최적화하기</a></li>
</ul>
<p class="caption"><span class="caption-text">오디오</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="audio_io_tutorial.html">Audio I/O</a></li>
<li class="toctree-l1"><a class="reference internal" href="audio_resampling_tutorial.html">Audio Resampling</a></li>
<li class="toctree-l1"><a class="reference internal" href="audio_data_augmentation_tutorial.html">Audio Data Augmentation</a></li>
<li class="toctree-l1"><a class="reference internal" href="audio_feature_extractions_tutorial.html">Audio Feature Extractions</a></li>
<li class="toctree-l1"><a class="reference internal" href="audio_feature_augmentation_tutorial.html">Audio Feature Augmentation</a></li>
<li class="toctree-l1"><a class="reference internal" href="audio_datasets_tutorial.html">Audio Datasets</a></li>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/speech_recognition_pipeline_tutorial.html">Speech Recognition with Wav2Vec2</a></li>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/speech_command_classification_with_torchaudio_tutorial.html">Speech Command Classification with torchaudio</a></li>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/text_to_speech_with_torchaudio.html">Text-to-speech with torchaudio</a></li>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/forced_alignment_with_torchaudio_tutorial.html">Forced Alignment with Wav2Vec2</a></li>
</ul>
<p class="caption"><span class="caption-text">텍스트</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="transformer_tutorial.html">nn.Transformer 와 TorchText 로 시퀀스-투-시퀀스(Sequence-to-Sequence) 모델링하기</a></li>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/char_rnn_classification_tutorial.html">기초부터 시작하는 NLP: 문자-단위 RNN으로 이름 분류하기</a></li>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/char_rnn_generation_tutorial.html">기초부터 시작하는 NLP:  문자-단위 RNN으로 이름 생성하기</a></li>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/seq2seq_translation_tutorial.html">기초부터 시작하는 NLP: Sequence to Sequence 네트워크와 Attention을 이용한 번역</a></li>
<li class="toctree-l1"><a class="reference internal" href="text_sentiment_ngrams_tutorial.html">torchtext 라이브러리로 텍스트 분류하기</a></li>
<li class="toctree-l1"><a class="reference internal" href="translation_transformer.html">nn.Transformer와 torchtext로 언어 번역하기</a></li>
</ul>
<p class="caption"><span class="caption-text">강화학습</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/reinforcement_q_learning.html">강화 학습 (DQN) 튜토리얼</a></li>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/mario_rl_tutorial.html">Train a Mario-playing RL Agent</a></li>
</ul>
<p class="caption"><span class="caption-text">PyTorch 모델을 프로덕션 환경에 배포하기</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/flask_rest_api_tutorial.html">Flask를 사용하여 Python에서 PyTorch를 REST API로 배포하기</a></li>
<li class="toctree-l1"><a class="reference internal" href="Intro_to_TorchScript_tutorial.html">TorchScript 소개</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/cpp_export.html">C++에서 TorchScript 모델 로딩하기</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/super_resolution_with_onnxruntime.html">(선택) PyTorch 모델을 ONNX으로 변환하고 ONNX 런타임에서 실행하기</a></li>
</ul>
<p class="caption"><span class="caption-text">Code Transforms with FX</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/fx_conv_bn_fuser.html">(베타) FX에서 합성곱/배치 정규화(Convolution/Batch Norm) 결합기(Fuser) 만들기</a></li>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/fx_profiling_tutorial.html">(beta) Building a Simple CPU Performance Profiler with FX</a></li>
</ul>
<p class="caption"><span class="caption-text">프론트엔드 API</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/memory_format_tutorial.html">(베타) PyTorch를 사용한 Channels Last 메모리 형식</a></li>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/forward_ad_usage.html">Forward-mode Automatic Differentiation (Beta)</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/cpp_frontend.html">PyTorch C++ 프론트엔드 사용하기</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/torch-script-parallelism.html">TorchScript의 동적 병렬 처리(Dynamic Parallelism)</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/cpp_autograd.html">C++ 프론트엔드의 자동 미분 (autograd)</a></li>
</ul>
<p class="caption"><span class="caption-text">PyTorch 확장하기</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/custom_function_double_backward_tutorial.html">Double Backward with Custom Functions</a></li>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/custom_function_conv_bn_tutorial.html">Fusing Convolution and Batch Norm using Custom Function</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/cpp_extension.html">Custom C++ and CUDA Extensions</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/torch_script_custom_ops.html">Extending TorchScript with Custom C++ Operators</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/torch_script_custom_classes.html">커스텀 C++ 클래스로 TorchScript 확장하기</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/dispatcher.html">Registering a Dispatched Operator in C++</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/extend_dispatcher.html">Extending dispatcher for a new backend in C++</a></li>
</ul>
<p class="caption"><span class="caption-text">모델 최적화</span></p>
<ul class="current">
<li class="toctree-l1"><a class="reference internal" href="profiler.html">PyTorch 모듈 프로파일링 하기</a></li>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/tensorboard_profiler_tutorial.html">PyTorch Profiler With TensorBoard</a></li>
<li class="toctree-l1"><a class="reference internal" href="hyperparameter_tuning_tutorial.html">Hyperparameter tuning with Ray Tune</a></li>
<li class="toctree-l1 current"><a class="current reference internal" href="#">배포를 위한 비전 트랜스포머(Vision Transformer) 모델 최적화하기</a></li>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/parametrizations.html">Parametrizations Tutorial</a></li>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/pruning_tutorial.html">가지치기 기법(Pruning) 튜토리얼</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/dynamic_quantization_tutorial.html">(베타) LSTM 기반 단어 단위 언어 모델의 동적 양자화</a></li>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/dynamic_quantization_bert_tutorial.html">(베타) BERT 모델 동적 양자화하기</a></li>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/quantized_transfer_learning_tutorial.html">(베타) 컴퓨터 비전 튜토리얼을 위한 양자화된 전이학습(Quantized Transfer Learning)</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/static_quantization_tutorial.html">(beta) Static Quantization with Eager Mode in PyTorch</a></li>
</ul>
<p class="caption"><span class="caption-text">병렬 및 분산 학습</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="dist_overview.html">PyTorch Distributed Overview</a></li>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/model_parallel_tutorial.html">단일 머신을 사용한 모델 병렬화 모범 사례</a></li>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/ddp_tutorial.html">분산 데이터 병렬 처리 시작하기</a></li>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/dist_tuto.html">PyTorch로 분산 어플리케이션 개발하기</a></li>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/rpc_tutorial.html">Getting Started with Distributed RPC Framework</a></li>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/rpc_param_server_tutorial.html">Implementing a Parameter Server Using Distributed RPC Framework</a></li>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/dist_pipeline_parallel_tutorial.html">Distributed Pipeline Parallelism Using RPC</a></li>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/rpc_async_execution.html">Implementing Batch RPC Processing Using Asynchronous Executions</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/rpc_ddp_tutorial.html">분산 데이터 병렬(DDP)과 분산 RPC 프레임워크 결합</a></li>
<li class="toctree-l1"><a class="reference internal" href="../intermediate/pipeline_tutorial.html">파이프라인 병렬화로 트랜스포머 모델 학습시키기</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/ddp_pipeline.html">분산 데이터 병렬 처리와 병렬 처리 파이프라인을 사용한 트랜스포머 모델 학습</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/generic_join.html">Distributed Training with Uneven Inputs Using the Join Context Manager</a></li>
</ul>
<p class="caption"><span class="caption-text">Mobile</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="deeplabv3_on_ios.html">iOS에서의 이미지 분할 DeepLapV3</a></li>
<li class="toctree-l1"><a class="reference internal" href="deeplabv3_on_android.html">안드로이드에서의 이미지 분할 DeepLapV3</a></li>
</ul>

            
        </div>
      </div>
    </nav>

    <div class="pytorch-container">
      <div class="pytorch-page-level-bar" id="pytorch-page-level-bar">
        <div class="pytorch-breadcrumbs-wrapper">
          

<div role="navigation" aria-label="breadcrumbs navigation">

  <ul class="pytorch-breadcrumbs">
    
      <li>
        <a href="../index.html">
          
            Tutorials
          
        </a> &gt;
      </li>

        
      <li>배포를 위한 비전 트랜스포머(Vision Transformer) 모델 최적화하기</li>
    
    
      <li class="pytorch-breadcrumbs-aside">
        
            
            <a href="../_sources/beginner/vt_tutorial.rst.txt" rel="nofollow"><img src="../_static/images/view-page-source-icon.svg"></a>
          
        
      </li>
    
  </ul>

  
</div>
        </div>

        <div class="pytorch-shortcuts-wrapper" id="pytorch-shortcuts-wrapper">
          Shortcuts
        </div>
      </div>

      <section data-toggle="wy-nav-shift" id="pytorch-content-wrap" class="pytorch-content-wrap">
        <div class="pytorch-content-left">

        
          <div class="pytorch-call-to-action-links">
            <div id="tutorial-type">beginner/vt_tutorial</div>

            <div id="google-colab-link">
              <img class="call-to-action-img" src="../_static/images/pytorch-colab.svg"/>
              <div class="call-to-action-desktop-view">Run in Google Colab</div>
              <div class="call-to-action-mobile-view">Colab</div>
            </div>
            <div id="download-notebook-link">
              <img class="call-to-action-notebook-img" src="../_static/images/pytorch-download.svg"/>
              <div class="call-to-action-desktop-view">Download Notebook</div>
              <div class="call-to-action-mobile-view">Notebook</div>
            </div>
            <div id="github-view-link">
              <img class="call-to-action-img" src="../_static/images/pytorch-github.svg"/>
              <div class="call-to-action-desktop-view">View on GitHub</div>
              <div class="call-to-action-mobile-view">GitHub</div>
            </div>
          </div>

        
          <div class="rst-content">
          
            <div role="main" class="main-content" itemscope="itemscope" itemtype="http://schema.org/Article">
             <article itemprop="articleBody" id="pytorch-article" class="pytorch-article">
              
  <div class="sphx-glr-download-link-note admonition note">
<p class="first admonition-title">Note</p>
<p class="last">Click <a class="reference internal" href="#sphx-glr-download-beginner-vt-tutorial-py"><span class="std std-ref">here</span></a> to download the full example code</p>
</div>
<div class="sphx-glr-example-title section" id="vision-transformer">
<span id="sphx-glr-beginner-vt-tutorial-py"></span><h1>배포를 위한 비전 트랜스포머(Vision Transformer) 모델 최적화하기<a class="headerlink" href="#vision-transformer" title="Permalink to this headline">¶</a></h1>
<p>Authors : <a class="reference external" href="https://github.com/jeffxtang">Jeff Tang</a>, <a class="reference external" href="https://github.com/gchauhan/">Geeta Chauhan</a>
번역 : <a class="reference external" href="https://github.com/Taeyoung96/">김태영</a></p>
<p>비전 트랜스포머(Vision Transformer)는 자연어 처리 분야에서 소개된
최고 수준의 결과를 달성한 최신의 어텐션 기반(attention-based) 트랜스포머 모델을
컴퓨터 비전 분야에 적용을 한 모델입니다.
FaceBook에서 발표한 Data-efficient Image Transformers는 <a class="reference external" href="https://ai.facebook.com/blog/data-efficient-image-transformers-a-promising-new-technique-for-image-classification">DeiT</a>
이미지 분류를 위해 ImageNet 데이터셋을 통해 훈련된
비전 트랜스포머 모델입니다.</p>
<p>이번 튜토리얼에서는, DeiT가 무엇인지 그리고 어떻게 사용하는지 다룰 것입니다.
그 다음 스크립팅, 양자화, 최적화, 그리고 iOS와 안드로이드 앱 안에서
모델을 사용하는 전체적인 단계를 수행해 볼 것입니다.
또한, 양자화와 최적화가 된 모델과 양자화와 최적화가 되지 않은 모델을 비교해 볼 것이며,
단계를 수행해 가면서 양자화와 최적화를 적용한 모델이 얼마나 이점을 가지는지 볼 것입니다.
DeiT란 무엇인가
——————–</p>
<p>합성곱 신경망(CNNs)은 2012년 딥러닝이 시작된 이후
이미지 분류를 수행할 때 주요한 모델이였습니다. 그러나 합성곱 신경망은 일반적으로
최첨단의 결과를 달성하기 위해 훈련에 수억 개의 이미지가 필요했습니다.
DeiT는 훈련에 더 적은 데이터와 컴퓨팅 자원을 필요로 하는 비전 트랜스포머 모델이며,
최신 CNN 모델과 이미지 분류를 수행하는데 경쟁을 합니다.
이는 DeiT의 두 가지 주요 구성 요소에 의해 가능하게 되었습니다.</p>
<ul class="simple">
<li>훨씬 더 큰 데이터 세트에 대한 훈련을 시뮬레이션하는 데이터 증강(augmentation)</li>
<li>트랜스포머 네트워크에 CNN의 출력값을 그대로 증류(distillation)하여 학습할 수 있도록 하는 기법</li>
</ul>
<p>DeiT는 제한된 데이터와 자원을 활용하여 컴퓨터 비전 태스크(task)에 트랜스포머 모델을
성공적으로 적용할 수 있음을 보여줍니다.
DeiT의 좀 더 자세한 내용을 원한다면, <a class="reference external" href="https://github.com/facebookresearch/deit">저장소</a>
와 <a class="reference external" href="https://arxiv.org/abs/2012.12877">논문</a> 을 참고하시길 바랍니다.</p>
<div class="section" id="id4">
<h2>DeiT를 활용한 이미지 분류<a class="headerlink" href="#id4" title="Permalink to this headline">¶</a></h2>
<p>DeiT를 사용하여 이미지를 분류하는 방법에 대한 자세한 정보는 DeiT 저장소에 README를 참고하시길 바랍니다.
빠른 테스트를 위해서, 먼저 필요한 패키지들을
설치합니다:</p>
<p>pip install torch torchvision timm pandas requests</p>
<p>Google Colab에서는 아래와 같이 실행합니다:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="c1"># !pip install timm pandas requests</span>
</pre></div>
</div>
<p>그런 다음 아래 스크립트를 실행합니다:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="kn">from</span> <span class="nn">PIL</span> <span class="kn">import</span> <span class="n">Image</span>
<span class="kn">import</span> <span class="nn">torch</span>
<span class="kn">import</span> <span class="nn">timm</span>
<span class="kn">import</span> <span class="nn">requests</span>
<span class="kn">import</span> <span class="nn">torchvision.transforms</span> <span class="k">as</span> <span class="nn">transforms</span>
<span class="kn">from</span> <span class="nn">timm.data.constants</span> <span class="kn">import</span> <span class="n">IMAGENET_DEFAULT_MEAN</span><span class="p">,</span> <span class="n">IMAGENET_DEFAULT_STD</span>

<span class="nb">print</span><span class="p">(</span><span class="n">torch</span><span class="o">.</span><span class="n">__version__</span><span class="p">)</span>
<span class="c1"># Pytorch 버전은 1.8.0 이어야 합니다.</span>


<span class="n">model</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">hub</span><span class="o">.</span><span class="n">load</span><span class="p">(</span><span class="s1">&#39;facebookresearch/deit:main&#39;</span><span class="p">,</span> <span class="s1">&#39;deit_base_patch16_224&#39;</span><span class="p">,</span> <span class="n">pretrained</span><span class="o">=</span><span class="kc">True</span><span class="p">)</span>
<span class="n">model</span><span class="o">.</span><span class="n">eval</span><span class="p">()</span>

<span class="n">transform</span> <span class="o">=</span> <span class="n">transforms</span><span class="o">.</span><span class="n">Compose</span><span class="p">([</span>
    <span class="n">transforms</span><span class="o">.</span><span class="n">Resize</span><span class="p">(</span><span class="mi">256</span><span class="p">,</span> <span class="n">interpolation</span><span class="o">=</span><span class="mi">3</span><span class="p">),</span>
    <span class="n">transforms</span><span class="o">.</span><span class="n">CenterCrop</span><span class="p">(</span><span class="mi">224</span><span class="p">),</span>
    <span class="n">transforms</span><span class="o">.</span><span class="n">ToTensor</span><span class="p">(),</span>
    <span class="n">transforms</span><span class="o">.</span><span class="n">Normalize</span><span class="p">(</span><span class="n">IMAGENET_DEFAULT_MEAN</span><span class="p">,</span> <span class="n">IMAGENET_DEFAULT_STD</span><span class="p">),</span>
<span class="p">])</span>

<span class="n">img</span> <span class="o">=</span> <span class="n">Image</span><span class="o">.</span><span class="n">open</span><span class="p">(</span><span class="n">requests</span><span class="o">.</span><span class="n">get</span><span class="p">(</span><span class="s2">&quot;https://raw.githubusercontent.com/pytorch/ios-demo-app/master/HelloWorld/HelloWorld/HelloWorld/image.png&quot;</span><span class="p">,</span> <span class="n">stream</span><span class="o">=</span><span class="kc">True</span><span class="p">)</span><span class="o">.</span><span class="n">raw</span><span class="p">)</span>
<span class="n">img</span> <span class="o">=</span> <span class="n">transform</span><span class="p">(</span><span class="n">img</span><span class="p">)[</span><span class="kc">None</span><span class="p">,]</span>
<span class="n">out</span> <span class="o">=</span> <span class="n">model</span><span class="p">(</span><span class="n">img</span><span class="p">)</span>
<span class="n">clsidx</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">argmax</span><span class="p">(</span><span class="n">out</span><span class="p">)</span>
<span class="nb">print</span><span class="p">(</span><span class="n">clsidx</span><span class="o">.</span><span class="n">item</span><span class="p">())</span>
</pre></div>
</div>
<p class="sphx-glr-script-out">Out:</p>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>1.10.0+cu102
269
</pre></div>
</div>
<p>ImageNet 목록에 따라 <a class="reference external" href="https://gist.github.com/yrevar/942d3a0ac09ec9e5eb3a">라벨(labels) 파일</a>
클래스 인덱스의 출력은 269여야 하며, 이는 ‘timber wolf, grey wolf, gray wolf, Canis lupus’에 매핑됩니다.</p>
<p>이제 DeiT 모델을 사용하여 이미지들을 분류할 수 있음을 확인했습니다.
iOS 및 Android 앱에서 실행할 수 있도록 모델을 수정하는 방법을 살펴보겠습니다.</p>
</div>
<div class="section" id="id5">
<h2>DeiT 스크립팅<a class="headerlink" href="#id5" title="Permalink to this headline">¶</a></h2>
<p>모바일에서 이 모델을 사용하려면, 우리는 첫번째로 모델 스크립팅이 필요합니다.
전체적인 개요는 <a class="reference external" href="https://tutorials.pytorch.kr/recipes/script_optimized.html">스크립트 그리고 최적화 레시피</a>
에서 확인할 수 있습니다. 아래 코드를 실행하여 이전 단계에서 사용한 DeiT 모델을
모바일에서 실행할 수 있는 TorchScript 형식으로 변환합니다.</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">model</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">hub</span><span class="o">.</span><span class="n">load</span><span class="p">(</span><span class="s1">&#39;facebookresearch/deit:main&#39;</span><span class="p">,</span> <span class="s1">&#39;deit_base_patch16_224&#39;</span><span class="p">,</span> <span class="n">pretrained</span><span class="o">=</span><span class="kc">True</span><span class="p">)</span>
<span class="n">model</span><span class="o">.</span><span class="n">eval</span><span class="p">()</span>
<span class="n">scripted_model</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">jit</span><span class="o">.</span><span class="n">script</span><span class="p">(</span><span class="n">model</span><span class="p">)</span>
<span class="n">scripted_model</span><span class="o">.</span><span class="n">save</span><span class="p">(</span><span class="s2">&quot;fbdeit_scripted.pt&quot;</span><span class="p">)</span>
</pre></div>
</div>
<p>약 346MB 크기의 스크립팅된 모델 파일 fbdeit_scripted.pt가 생성됩니다.</p>
</div>
<div class="section" id="id7">
<h2>DeiT 양자화<a class="headerlink" href="#id7" title="Permalink to this headline">¶</a></h2>
<p>추론 정확도를 거의 동일하게 유지하면서 훈련된 모델 크기를 크게 줄이기 위해
모델에 양자화를 적용할 수 있습니다.
DeiT에서 사용된 트랜스포머 모델 덕분에,
모델에 동적 양자화를 쉽게 적용할 수 있습니다.
왜나하면 동적 양자화는 LSTM 모델과 트랜스포머 모델에서 가장 잘 적용되기 때문입니다.
(자세한 내용은 <a class="reference external" href="https://pytorch.org/docs/stable/quantization.html?highlight=quantization#dynamic-quantization">여기</a>
를 참고하세요.)</p>
<p>아래의 코드를 실행시켜 봅시다.</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="c1"># 서버 추론을 위해 &#39;fbgemm&#39;을, 모바일 추론을 위해 &#39;qnnpack&#39;을 사용해 봅시다.</span>
<span class="n">backend</span> <span class="o">=</span> <span class="s2">&quot;fbgemm&quot;</span> <span class="c1"># 이 주피터 노트북에서는 양자화된 모델의 더 느린 추론 속도를 일으키는 qnnpack으로 대체되었습니다.</span>
<span class="n">model</span><span class="o">.</span><span class="n">qconfig</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">quantization</span><span class="o">.</span><span class="n">get_default_qconfig</span><span class="p">(</span><span class="n">backend</span><span class="p">)</span>
<span class="n">torch</span><span class="o">.</span><span class="n">backends</span><span class="o">.</span><span class="n">quantized</span><span class="o">.</span><span class="n">engine</span> <span class="o">=</span> <span class="n">backend</span>

<span class="n">quantized_model</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">quantization</span><span class="o">.</span><span class="n">quantize_dynamic</span><span class="p">(</span><span class="n">model</span><span class="p">,</span> <span class="n">qconfig_spec</span><span class="o">=</span><span class="p">{</span><span class="n">torch</span><span class="o">.</span><span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">},</span> <span class="n">dtype</span><span class="o">=</span><span class="n">torch</span><span class="o">.</span><span class="n">qint8</span><span class="p">)</span>
<span class="n">scripted_quantized_model</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">jit</span><span class="o">.</span><span class="n">script</span><span class="p">(</span><span class="n">quantized_model</span><span class="p">)</span>
<span class="n">scripted_quantized_model</span><span class="o">.</span><span class="n">save</span><span class="p">(</span><span class="s2">&quot;fbdeit_scripted_quantized.pt&quot;</span><span class="p">)</span>
</pre></div>
</div>
<p>fbdeit_quantized_scripted.pt 모델의 스크립팅과 양자화가 적용된 버전이 만들어졌습니다.
모델의 크기는 단지 89MB 입니다.
양자화가 적용되지 않은 모델의 크기인 346MB보다 74%나 감소했습니다!</p>
<p>동일한 추론 결과를 만들기 위해 <a href="#id9"><span class="problematic" id="id10">``</span></a>scripted_quantized_model``을
사용해 봅시다.</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">out</span> <span class="o">=</span> <span class="n">scripted_quantized_model</span><span class="p">(</span><span class="n">img</span><span class="p">)</span>
<span class="n">clsidx</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">argmax</span><span class="p">(</span><span class="n">out</span><span class="p">)</span>
<span class="nb">print</span><span class="p">(</span><span class="n">clsidx</span><span class="o">.</span><span class="n">item</span><span class="p">())</span>
<span class="c1"># 동일한 출력 결과인 269가 출력 되어야 합니다.</span>
</pre></div>
</div>
<p class="sphx-glr-script-out">Out:</p>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>269
</pre></div>
</div>
</div>
<div class="section" id="id11">
<h2>DeiT 최적화<a class="headerlink" href="#id11" title="Permalink to this headline">¶</a></h2>
<p>모바일에 스크립트 되고 양자화된 모델을 사용하기 위한
마지막 단계는 최적화입니다.</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="kn">from</span> <span class="nn">torch.utils.mobile_optimizer</span> <span class="kn">import</span> <span class="n">optimize_for_mobile</span>
<span class="n">optimized_scripted_quantized_model</span> <span class="o">=</span> <span class="n">optimize_for_mobile</span><span class="p">(</span><span class="n">scripted_quantized_model</span><span class="p">)</span>
<span class="n">optimized_scripted_quantized_model</span><span class="o">.</span><span class="n">save</span><span class="p">(</span><span class="s2">&quot;fbdeit_optimized_scripted_quantized.pt&quot;</span><span class="p">)</span>
</pre></div>
</div>
<p>생성된 fbdeit_optimized_scripted_quantized.pt 파일은
양자화되고 스크립트되지만 최적화되지 않은 모델과 크기가 거의 같습니다.
추론 결과는 동일하게 유지됩니다.</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">out</span> <span class="o">=</span> <span class="n">optimized_scripted_quantized_model</span><span class="p">(</span><span class="n">img</span><span class="p">)</span>
<span class="n">clsidx</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">argmax</span><span class="p">(</span><span class="n">out</span><span class="p">)</span>
<span class="nb">print</span><span class="p">(</span><span class="n">clsidx</span><span class="o">.</span><span class="n">item</span><span class="p">())</span>
<span class="c1"># 다시 한번, 동일한 출력 결과인 269가 출력 되어야 합니다.</span>
</pre></div>
</div>
<p class="sphx-glr-script-out">Out:</p>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>269
</pre></div>
</div>
</div>
<div class="section" id="lite-interpreter">
<h2>라이트 인터프리터(Lite interpreter) 사용<a class="headerlink" href="#lite-interpreter" title="Permalink to this headline">¶</a></h2>
<p>라이트 인터프리터를 사용하면 얼마나 모델의 사이즈가 작아지고, 추론 시간이 짧아지는지
결과를 확인해 봅시다. 이제 좀 더 가벼운 버전의 모델을 만들어 봅시다.</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">optimized_scripted_quantized_model</span><span class="o">.</span><span class="n">_save_for_lite_interpreter</span><span class="p">(</span><span class="s2">&quot;fbdeit_optimized_scripted_quantized_lite.ptl&quot;</span><span class="p">)</span>
<span class="n">ptl</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">jit</span><span class="o">.</span><span class="n">load</span><span class="p">(</span><span class="s2">&quot;fbdeit_optimized_scripted_quantized_lite.ptl&quot;</span><span class="p">)</span>
</pre></div>
</div>
<p>가벼운 모델의 크기는 그렇지 않은 버전의 모델 크기와 비슷하지만,
모바일에서 가벼운 버전을 실행하면 추론 속도가 빨라질 것으로 예상됩니다.</p>
</div>
<div class="section" id="id12">
<h2>추론 속도 비교<a class="headerlink" href="#id12" title="Permalink to this headline">¶</a></h2>
<p>네 가지 모델(원본 모델, 스크립트된 모델, 스크립트와 양자화를 적용한 모델,
스크립트와 양자화를 적용한 후 최적화한 모델)의 추론 속도가 어떻게 다른지 확인해 봅시다.</p>
<p>아래의 코드를 실행해 봅시다.</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="k">with</span> <span class="n">torch</span><span class="o">.</span><span class="n">autograd</span><span class="o">.</span><span class="n">profiler</span><span class="o">.</span><span class="n">profile</span><span class="p">(</span><span class="n">use_cuda</span><span class="o">=</span><span class="kc">False</span><span class="p">)</span> <span class="k">as</span> <span class="n">prof1</span><span class="p">:</span>
    <span class="n">out</span> <span class="o">=</span> <span class="n">model</span><span class="p">(</span><span class="n">img</span><span class="p">)</span>
<span class="k">with</span> <span class="n">torch</span><span class="o">.</span><span class="n">autograd</span><span class="o">.</span><span class="n">profiler</span><span class="o">.</span><span class="n">profile</span><span class="p">(</span><span class="n">use_cuda</span><span class="o">=</span><span class="kc">False</span><span class="p">)</span> <span class="k">as</span> <span class="n">prof2</span><span class="p">:</span>
    <span class="n">out</span> <span class="o">=</span> <span class="n">scripted_model</span><span class="p">(</span><span class="n">img</span><span class="p">)</span>
<span class="k">with</span> <span class="n">torch</span><span class="o">.</span><span class="n">autograd</span><span class="o">.</span><span class="n">profiler</span><span class="o">.</span><span class="n">profile</span><span class="p">(</span><span class="n">use_cuda</span><span class="o">=</span><span class="kc">False</span><span class="p">)</span> <span class="k">as</span> <span class="n">prof3</span><span class="p">:</span>
    <span class="n">out</span> <span class="o">=</span> <span class="n">scripted_quantized_model</span><span class="p">(</span><span class="n">img</span><span class="p">)</span>
<span class="k">with</span> <span class="n">torch</span><span class="o">.</span><span class="n">autograd</span><span class="o">.</span><span class="n">profiler</span><span class="o">.</span><span class="n">profile</span><span class="p">(</span><span class="n">use_cuda</span><span class="o">=</span><span class="kc">False</span><span class="p">)</span> <span class="k">as</span> <span class="n">prof4</span><span class="p">:</span>
    <span class="n">out</span> <span class="o">=</span> <span class="n">optimized_scripted_quantized_model</span><span class="p">(</span><span class="n">img</span><span class="p">)</span>
<span class="k">with</span> <span class="n">torch</span><span class="o">.</span><span class="n">autograd</span><span class="o">.</span><span class="n">profiler</span><span class="o">.</span><span class="n">profile</span><span class="p">(</span><span class="n">use_cuda</span><span class="o">=</span><span class="kc">False</span><span class="p">)</span> <span class="k">as</span> <span class="n">prof5</span><span class="p">:</span>
    <span class="n">out</span> <span class="o">=</span> <span class="n">ptl</span><span class="p">(</span><span class="n">img</span><span class="p">)</span>

<span class="nb">print</span><span class="p">(</span><span class="s2">&quot;original model: </span><span class="si">{:.2f}</span><span class="s2">ms&quot;</span><span class="o">.</span><span class="n">format</span><span class="p">(</span><span class="n">prof1</span><span class="o">.</span><span class="n">self_cpu_time_total</span><span class="o">/</span><span class="mi">1000</span><span class="p">))</span>
<span class="nb">print</span><span class="p">(</span><span class="s2">&quot;scripted model: </span><span class="si">{:.2f}</span><span class="s2">ms&quot;</span><span class="o">.</span><span class="n">format</span><span class="p">(</span><span class="n">prof2</span><span class="o">.</span><span class="n">self_cpu_time_total</span><span class="o">/</span><span class="mi">1000</span><span class="p">))</span>
<span class="nb">print</span><span class="p">(</span><span class="s2">&quot;scripted &amp; quantized model: </span><span class="si">{:.2f}</span><span class="s2">ms&quot;</span><span class="o">.</span><span class="n">format</span><span class="p">(</span><span class="n">prof3</span><span class="o">.</span><span class="n">self_cpu_time_total</span><span class="o">/</span><span class="mi">1000</span><span class="p">))</span>
<span class="nb">print</span><span class="p">(</span><span class="s2">&quot;scripted &amp; quantized &amp; optimized model: </span><span class="si">{:.2f}</span><span class="s2">ms&quot;</span><span class="o">.</span><span class="n">format</span><span class="p">(</span><span class="n">prof4</span><span class="o">.</span><span class="n">self_cpu_time_total</span><span class="o">/</span><span class="mi">1000</span><span class="p">))</span>
<span class="nb">print</span><span class="p">(</span><span class="s2">&quot;lite model: </span><span class="si">{:.2f}</span><span class="s2">ms&quot;</span><span class="o">.</span><span class="n">format</span><span class="p">(</span><span class="n">prof5</span><span class="o">.</span><span class="n">self_cpu_time_total</span><span class="o">/</span><span class="mi">1000</span><span class="p">))</span>
</pre></div>
</div>
<p class="sphx-glr-script-out">Out:</p>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>original model: 2101.55ms
scripted model: 2032.05ms
scripted &amp; quantized model: 1739.19ms
scripted &amp; quantized &amp; optimized model: 1824.82ms
lite model: 1804.25ms
</pre></div>
</div>
<p>Google Colab에서 실행 시킨 결과는 다음과 같습니다.</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">original</span> <span class="n">model</span><span class="p">:</span> <span class="mf">1236.69</span><span class="n">ms</span>
<span class="n">scripted</span> <span class="n">model</span><span class="p">:</span> <span class="mf">1226.72</span><span class="n">ms</span>
<span class="n">scripted</span> <span class="o">&amp;</span> <span class="n">quantized</span> <span class="n">model</span><span class="p">:</span> <span class="mf">593.19</span><span class="n">ms</span>
<span class="n">scripted</span> <span class="o">&amp;</span> <span class="n">quantized</span> <span class="o">&amp;</span> <span class="n">optimized</span> <span class="n">model</span><span class="p">:</span> <span class="mf">598.01</span><span class="n">ms</span>
<span class="n">lite</span> <span class="n">model</span><span class="p">:</span> <span class="mf">600.72</span><span class="n">ms</span>
</pre></div>
</div>
<p>다음 결과는 각 모델이 소요한 추론 시간과
원본 모델에 대한 각 모델의 감소율을 요약한 것입니다.</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="kn">import</span> <span class="nn">pandas</span> <span class="k">as</span> <span class="nn">pd</span>
<span class="kn">import</span> <span class="nn">numpy</span> <span class="k">as</span> <span class="nn">np</span>

<span class="n">df</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">DataFrame</span><span class="p">({</span><span class="s1">&#39;Model&#39;</span><span class="p">:</span> <span class="p">[</span><span class="s1">&#39;original model&#39;</span><span class="p">,</span><span class="s1">&#39;scripted model&#39;</span><span class="p">,</span> <span class="s1">&#39;scripted &amp; quantized model&#39;</span><span class="p">,</span> <span class="s1">&#39;scripted &amp; quantized &amp; optimized model&#39;</span><span class="p">,</span> <span class="s1">&#39;lite model&#39;</span><span class="p">]})</span>
<span class="n">df</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">concat</span><span class="p">([</span><span class="n">df</span><span class="p">,</span> <span class="n">pd</span><span class="o">.</span><span class="n">DataFrame</span><span class="p">([</span>
    <span class="p">[</span><span class="s2">&quot;</span><span class="si">{:.2f}</span><span class="s2">ms&quot;</span><span class="o">.</span><span class="n">format</span><span class="p">(</span><span class="n">prof1</span><span class="o">.</span><span class="n">self_cpu_time_total</span><span class="o">/</span><span class="mi">1000</span><span class="p">),</span> <span class="s2">&quot;0%&quot;</span><span class="p">],</span>
    <span class="p">[</span><span class="s2">&quot;</span><span class="si">{:.2f}</span><span class="s2">ms&quot;</span><span class="o">.</span><span class="n">format</span><span class="p">(</span><span class="n">prof2</span><span class="o">.</span><span class="n">self_cpu_time_total</span><span class="o">/</span><span class="mi">1000</span><span class="p">),</span>
     <span class="s2">&quot;</span><span class="si">{:.2f}</span><span class="s2">%&quot;</span><span class="o">.</span><span class="n">format</span><span class="p">((</span><span class="n">prof1</span><span class="o">.</span><span class="n">self_cpu_time_total</span><span class="o">-</span><span class="n">prof2</span><span class="o">.</span><span class="n">self_cpu_time_total</span><span class="p">)</span><span class="o">/</span><span class="n">prof1</span><span class="o">.</span><span class="n">self_cpu_time_total</span><span class="o">*</span><span class="mi">100</span><span class="p">)],</span>
    <span class="p">[</span><span class="s2">&quot;</span><span class="si">{:.2f}</span><span class="s2">ms&quot;</span><span class="o">.</span><span class="n">format</span><span class="p">(</span><span class="n">prof3</span><span class="o">.</span><span class="n">self_cpu_time_total</span><span class="o">/</span><span class="mi">1000</span><span class="p">),</span>
     <span class="s2">&quot;</span><span class="si">{:.2f}</span><span class="s2">%&quot;</span><span class="o">.</span><span class="n">format</span><span class="p">((</span><span class="n">prof1</span><span class="o">.</span><span class="n">self_cpu_time_total</span><span class="o">-</span><span class="n">prof3</span><span class="o">.</span><span class="n">self_cpu_time_total</span><span class="p">)</span><span class="o">/</span><span class="n">prof1</span><span class="o">.</span><span class="n">self_cpu_time_total</span><span class="o">*</span><span class="mi">100</span><span class="p">)],</span>
    <span class="p">[</span><span class="s2">&quot;</span><span class="si">{:.2f}</span><span class="s2">ms&quot;</span><span class="o">.</span><span class="n">format</span><span class="p">(</span><span class="n">prof4</span><span class="o">.</span><span class="n">self_cpu_time_total</span><span class="o">/</span><span class="mi">1000</span><span class="p">),</span>
     <span class="s2">&quot;</span><span class="si">{:.2f}</span><span class="s2">%&quot;</span><span class="o">.</span><span class="n">format</span><span class="p">((</span><span class="n">prof1</span><span class="o">.</span><span class="n">self_cpu_time_total</span><span class="o">-</span><span class="n">prof4</span><span class="o">.</span><span class="n">self_cpu_time_total</span><span class="p">)</span><span class="o">/</span><span class="n">prof1</span><span class="o">.</span><span class="n">self_cpu_time_total</span><span class="o">*</span><span class="mi">100</span><span class="p">)],</span>
    <span class="p">[</span><span class="s2">&quot;</span><span class="si">{:.2f}</span><span class="s2">ms&quot;</span><span class="o">.</span><span class="n">format</span><span class="p">(</span><span class="n">prof5</span><span class="o">.</span><span class="n">self_cpu_time_total</span><span class="o">/</span><span class="mi">1000</span><span class="p">),</span>
     <span class="s2">&quot;</span><span class="si">{:.2f}</span><span class="s2">%&quot;</span><span class="o">.</span><span class="n">format</span><span class="p">((</span><span class="n">prof1</span><span class="o">.</span><span class="n">self_cpu_time_total</span><span class="o">-</span><span class="n">prof5</span><span class="o">.</span><span class="n">self_cpu_time_total</span><span class="p">)</span><span class="o">/</span><span class="n">prof1</span><span class="o">.</span><span class="n">self_cpu_time_total</span><span class="o">*</span><span class="mi">100</span><span class="p">)]],</span>
    <span class="n">columns</span><span class="o">=</span><span class="p">[</span><span class="s1">&#39;Inference Time&#39;</span><span class="p">,</span> <span class="s1">&#39;Reduction&#39;</span><span class="p">])],</span> <span class="n">axis</span><span class="o">=</span><span class="mi">1</span><span class="p">)</span>

<span class="nb">print</span><span class="p">(</span><span class="n">df</span><span class="p">)</span>

<span class="sd">&quot;&quot;&quot;</span>
<span class="sd">        Model                             Inference Time    Reduction</span>
<span class="sd">0   original model                             1236.69ms           0%</span>
<span class="sd">1   scripted model                             1226.72ms        0.81%</span>
<span class="sd">2   scripted &amp; quantized model                  593.19ms       52.03%</span>
<span class="sd">3   scripted &amp; quantized &amp; optimized model      598.01ms       51.64%</span>
<span class="sd">4   lite model                                  600.72ms       51.43%</span>
<span class="sd">&quot;&quot;&quot;</span>
</pre></div>
</div>
<p class="sphx-glr-script-out">Out:</p>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Model Inference Time Reduction
0                          original model      2101.55ms        0%
1                          scripted model      2032.05ms     3.31%
2              scripted &amp; quantized model      1739.19ms    17.24%
3  scripted &amp; quantized &amp; optimized model      1824.82ms    13.17%
4                              lite model      1804.25ms    14.15%
</pre></div>
</div>
<div class="section" id="id13">
<h3>더 읽을거리<a class="headerlink" href="#id13" title="Permalink to this headline">¶</a></h3>
<ul class="simple">
<li><a class="reference external" href="https://ai.facebook.com/blog/data-efficient-image-transformers-a-promising-new-technique-for-image-classification">Facebook Data-efficient Image Transformers</a></li>
<li><a class="reference external" href="https://github.com/pytorch/ios-demo-app/tree/master/ViT4MNIST">Vision Transformer with ImageNet and MNIST on iOS</a></li>
<li><a class="reference external" href="https://github.com/pytorch/android-demo-app/tree/master/ViT4MNIST">Vision Transformer with ImageNet and MNIST on Android</a></li>
</ul>
<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 0 minutes  44.915 seconds)</p>
<div class="sphx-glr-footer class sphx-glr-footer-example docutils container" id="sphx-glr-download-beginner-vt-tutorial-py">
<div class="sphx-glr-download docutils container">
<a class="reference download internal" download="" href="../_downloads/ebdb4d03df673d81802813f1cdc96c81/vt_tutorial.py"><code class="xref download docutils literal notranslate"><span class="pre">Download</span> <span class="pre">Python</span> <span class="pre">source</span> <span class="pre">code:</span> <span class="pre">vt_tutorial.py</span></code></a></div>
<div class="sphx-glr-download docutils container">
<a class="reference download internal" download="" href="../_downloads/f1d2bc04941499e61ed06a393ff4db64/vt_tutorial.ipynb"><code class="xref download docutils literal notranslate"><span class="pre">Download</span> <span class="pre">Jupyter</span> <span class="pre">notebook:</span> <span class="pre">vt_tutorial.ipynb</span></code></a></div>
</div>
<p class="sphx-glr-signature"><a class="reference external" href="https://sphinx-gallery.readthedocs.io">Gallery generated by Sphinx-Gallery</a></p>
</div>
</div>
</div>


             </article>
             
            </div>
            <footer>
  
    <div class="rst-footer-buttons" role="navigation" aria-label="footer navigation">
      
        <a href="audio_io_tutorial.html" class="btn btn-neutral float-right" title="Audio I/O" accesskey="n" rel="next">Next <img src="../_static/images/chevron-right-orange.svg" class="next-page"></a>
      
      
        <a href="dcgan_faces_tutorial.html" class="btn btn-neutral" title="DCGAN 튜토리얼" accesskey="p" rel="prev"><img src="../_static/images/chevron-right-orange.svg" class="previous-page"> Previous</a>
      
    </div>
  

    <hr class="rating-hr hr-top">
      <div class="rating-container">
        <div class="rating-prompt">Rate this Tutorial</div>
        <div class="stars-outer">
          <i class="far fa-star" title="1 Star" data-behavior="tutorial-rating" data-count="1"></i>
          <i class="far fa-star" title="2 Stars" data-behavior="tutorial-rating" data-count="2"></i>
          <i class="far fa-star" title="3 Stars" data-behavior="tutorial-rating" data-count="3"></i>
          <i class="far fa-star" title="4 Stars" data-behavior="tutorial-rating" data-count="4"></i>
          <i class="far fa-star" title="5 Stars" data-behavior="tutorial-rating" data-count="5"></i>
        </div>
      </div>
    <hr class="rating-hr hr-bottom"/>

  
  <div role="contentinfo">
    <p>
        &copy; Copyright 2021, PyTorch &amp; PyTorch Korea Community.

    </p>
  </div>
    
      <div>
        Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a <a href="https://github.com/rtfd/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>.
      </div>
     

</footer>

          </div>
        </div>

        <div class="pytorch-content-right" id="pytorch-content-right">
          <div class="pytorch-right-menu" id="pytorch-right-menu">
            <div class="pytorch-side-scroll" id="pytorch-side-scroll-right">
              <ul>
<li><a class="reference internal" href="#">배포를 위한 비전 트랜스포머(Vision Transformer) 모델 최적화하기</a><ul>
<li><a class="reference internal" href="#id4">DeiT를 활용한 이미지 분류</a></li>
<li><a class="reference internal" href="#id5">DeiT 스크립팅</a></li>
<li><a class="reference internal" href="#id7">DeiT 양자화</a></li>
<li><a class="reference internal" href="#id11">DeiT 최적화</a></li>
<li><a class="reference internal" href="#lite-interpreter">라이트 인터프리터(Lite interpreter) 사용</a></li>
<li><a class="reference internal" href="#id12">추론 속도 비교</a><ul>
<li><a class="reference internal" href="#id13">더 읽을거리</a></li>
</ul>
</li>
</ul>
</li>
</ul>

            </div>
          </div>
        </div>
      </section>
    </div>

  
       <script type="text/javascript" id="documentation_options" data-url_root="../" src="../_static/documentation_options.js"></script>
         <script type="text/javascript" src="../_static/jquery.js"></script>
         <script type="text/javascript" src="../_static/underscore.js"></script>
         <script type="text/javascript" src="../_static/doctools.js"></script>
         <script type="text/javascript" src="../_static/clipboard.min.js"></script>
         <script type="text/javascript" src="../_static/copybutton.js"></script>
         <script type="text/javascript" src="https://cdn.jsdelivr.net/npm/katex@0.13.11/dist/katex.min.js"></script>
         <script type="text/javascript" src="https://cdn.jsdelivr.net/npm/katex@0.13.11/dist/contrib/auto-render.min.js"></script>
         <script type="text/javascript" src="../_static/katex_autorenderer.js"></script>
     

  <script type="text/javascript" src="../_static/js/vendor/popper.min.js"></script>
  <script type="text/javascript" src="../_static/js/vendor/bootstrap.min.js"></script>
  <script src="https://cdnjs.cloudflare.com/ajax/libs/list.js/1.5.0/list.min.js"></script>
  <script type="text/javascript" src="../_static/js/theme.js"></script>

  <script type="text/javascript">
      jQuery(function () {
          SphinxRtdTheme.Navigation.enable(true);
      });
  </script>
 

<script>
//add microsoft link

if(window.location.href.indexOf("/beginner/basics/")!= -1)
{
  var url="https://docs.microsoft.com/learn/paths/pytorch-fundamentals/?wt.mc_id=aiml-7486-cxa";
  switch(window.location.pathname.split("/").pop().replace('.html',''))
  {
    case"quickstart_tutorial":
      url="https://docs.microsoft.com/learn/modules/intro-machine-learning-pytorch/9-quickstart?WT.mc_id=aiml-7486-cxa";
      break;
    case"tensorqs_tutorial":
      url="https://docs.microsoft.com/learn/modules/intro-machine-learning-pytorch/2-tensors?WT.mc_id=aiml-7486-cxa";
      break;
    case"data_tutorial":
      url="https://docs.microsoft.com/learn/modules/intro-machine-learning-pytorch/3-data?WT.mc_id=aiml-7486-cxa";
      break;
    case"transforms_tutorial":
      url="https://docs.microsoft.com/learn/modules/intro-machine-learning-pytorch/4-transforms?WT.mc_id=aiml-7486-cxa";
      break;
    case"buildmodel_tutorial":
      url="https://docs.microsoft.com/learn/modules/intro-machine-learning-pytorch/5-model?WT.mc_id=aiml-7486-cxa";
      break;
    case"autogradqs_tutorial":
      url="https://docs.microsoft.com/learn/modules/intro-machine-learning-pytorch/6-autograd?WT.mc_id=aiml-7486-cxa";
      break;
    case"optimization_tutorial":
      url="https://docs.microsoft.com/learn/modules/intro-machine-learning-pytorch/7-optimization?WT.mc_id=aiml-7486-cxa";
      break;
    case"saveloadrun_tutorial":
      url="https://docs.microsoft.com/learn/modules/intro-machine-learning-pytorch/8-inference?WT.mc_id=aiml-7486-cxa";
    }

    $(".pytorch-call-to-action-links").children().first().before("<a href="+url+' data-behavior="call-to-action-event" data-response="Run in Microsoft Learn" target="_blank"><div id="microsoft-learn-link" style="padding-bottom: 0.625rem;border-bottom: 1px solid #f3f4f7;padding-right: 2.5rem;display: -webkit-box;  display: -ms-flexbox; isplay: flex; -webkit-box-align: center;-ms-flex-align: center;align-items: center;"><img class="call-to-action-img" src="../../_static/images/microsoft-logo.svg"/><div class="call-to-action-desktop-view">Run in Microsoft Learn</div><div class="call-to-action-mobile-view">Learn</div></div></a>')
  }
</script>

<script async src="https://www.googletagmanager.com/gtag/js?id=UA-71919972-3"></script>
<script>
  window.dataLayer = window.dataLayer || [];
  function gtag(){dataLayer.push(arguments);}
  gtag('js', new Date());
  gtag('config', 'UA-71919972-3');
</script>


<script>
  $("[data-behavior='call-to-action-event']").on('click', function(){
    ga('send', {
      hitType: 'event',
      eventCategory: $(this).attr("data-response"),
      eventAction: 'click',
      eventLabel: window.location.href
    });

    gtag('event', 'click', {
      'event_category': $(this).attr("data-response"),
      'event_label': $("h1").first().text(),
      'tutorial_link': window.location.href
    });
   });

   $("[data-behavior='tutorial-rating']").on('click', function(){
    gtag('event', 'click', {
      'event_category': 'Tutorial Rating',
      'event_label': $("h1").first().text(),
      'value': $(this).attr("data-count")
    });
   });

   if (location.pathname == "/") {
     $(".rating-container").hide();
     $(".hr-bottom").hide();
   }
</script>

<script type="text/javascript">
  var collapsedSections = ['파이토치(PyTorch) 레시피', '파이토치(PyTorch) 배우기', '이미지/비디오', '오디오', '텍스트', '강화학습', 'PyTorch 모델을 프로덕션 환경에 배포하기', 'Code Transforms with FX', '프론트엔드 API', 'PyTorch 확장하기', '모델 최적화', '병렬 및 분산 학습', 'Mobile'];
</script>


  <!-- Begin Footer -->

  <div class="container-fluid docs-tutorials-resources" id="docs-tutorials-resources">
    <div class="container">
      <div class="row">
        <div class="col-md-4 text-center">
          <h2>공식 문서 (영어)</h2>
          <p>PyTorch 공식 문서입니다.</p>
          <a id="orgTutorialLink" class="with-right-arrow" href="https://pytorch.org/docs/stable/index.html" target="_blank">공식 문서로 이동</a>
        </div>

        <div class="col-md-4 text-center">
          <h2>한국어 튜토리얼</h2>
          <p>한국어로 번역 중인 PyTorch 튜토리얼입니다.</p>
          <a class="with-right-arrow" href="https://tutorials.pytorch.kr">튜토리얼로 이동</a>
        </div>

        <div class="col-md-4 text-center">
          <h2>커뮤니티</h2>
          <p>다른 사용자들과 의견을 나눠보세요!</p>
          <a class="with-right-arrow" href="https://discuss.pytorch.kr">커뮤니티로 이동</a>
        </div>
      </div>
    </div>
  </div>

  <footer class="site-footer">
    <div class="container footer-container">
      <div class="footer-logo-wrapper">
        <a href="https://pytorch.kr/" class="footer-logo"></a>
      </div>

      <div class="footer-links-wrapper">
        <div class="footer-links-col">
          <ul>
            <li class="list-title"><a href="https://pytorch.org">PyTorch 홈페이지 (공식)</a></li>
            <li><a href="https://pytorch.org" target="_blank">공식 홈페이지</a></li>
            <li><a href="https://pytorch.org/tutorials" target="_blank">공식 튜토리얼</a></li>
            <li><a href="https://pytorch.org/docs" target="_blank">공식 문서</a></li>
          </ul>
        </div>

        <div class="footer-links-col">
          <ul>
            <li class="list-title"><a href="">한국 사용자 모임</a></li>
            <li><a href="https://pytorch.kr/about">사이트 소개</a></li>
            <li><a href="https://tutorials.pytorch.kr/">한국어 튜토리얼</a></li>
            <li><a href="https://github.com/9bow/PyTorch-tutorials-kr" target="_blank">한국어 튜토리얼 저장소</a></li>
          </ul>
        </div>

      </div>
      <div class="trademark-disclaimer">
        <ul>
          <li>이 사이트는 PyTorch 한국 사용자 커뮤니티로 Facebook, Inc에서 운영하는 사이트가 아닙니다. PyTorch, PyTorch 로고 및 모든 관련 표기는 Facebook, Inc의 상표입니다.</li>
          <li>This site is a user community and is not operated by Facebook, Inc. PyTorch, the PyTorch logo and any related marks are trademarks of Facebook, Inc.</li>
        </ul>
      </div>
    </div>
  </footer>

  <!-- End Footer -->

  <!-- Begin Mobile Menu -->

  <div class="mobile-main-menu">
    <div class="container-fluid">
      <div class="container">
        <div class="mobile-main-menu-header-container">
          <a class="header-logo" href="https://pytorch.kr/" aria-label="PyTorch"></a>
          <a class="main-menu-close-button" href="#" data-behavior="close-mobile-menu"></a>
        </div>
      </div>
    </div>

    <div class="mobile-main-menu-links-container">
      <div class="main-menu">
        <ul>
          <li>
            <a href="https://pytorch.kr/get-started">시작하기</a>
          </li>

          <li class="active">
            <a href="https://tutorials.pytorch.kr">튜토리얼</a>
          </li>

          <li>
            <a href="https://pytorch.kr/hub">허브</a>
          </li>

          <li>
            <a href="https://discuss.pytorch.kr">커뮤니티</a>
          </li>
        </ul>
      </div>
    </div>
  </div>

  <!-- End Mobile Menu -->

  <script type="text/javascript" src="../_static/js/vendor/anchor.min.js"></script>

  <script type="text/javascript">
    $(document).ready(function() {
      mobileMenu.bind();
      mobileTOC.bind();
      pytorchAnchors.bind();
      sideMenus.bind();
      scrollToAnchor.bind();
      highlightNavigation.bind();
      mainMenuDropdown.bind();
      filterTags.bind();

      // Add class to links that have code blocks, since we cannot create links in code blocks
      $("article.pytorch-article a span.pre").each(function(e) {
        $(this).closest("a").addClass("has-code");
      });
    })
  </script>
</body>
</html>