<!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>Introduction to PyTorch — 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" />
<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>
<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"><a class="reference internal" href="../vt_tutorial.html">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>
<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"><a class="reference internal" href="../vt_tutorial.html">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> >
</li>
<li>Introduction to PyTorch</li>
<li class="pytorch-breadcrumbs-aside">
<a href="../../_sources/beginner/nlp/pytorch_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/nlp/pytorch_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-nlp-pytorch-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="introduction-to-pytorch">
<span id="sphx-glr-beginner-nlp-pytorch-tutorial-py"></span><h1>Introduction to PyTorch<a class="headerlink" href="#introduction-to-pytorch" title="Permalink to this headline">¶</a></h1>
<div class="section" id="introduction-to-torch-s-tensor-library">
<h2>Introduction to Torch’s tensor library<a class="headerlink" href="#introduction-to-torch-s-tensor-library" title="Permalink to this headline">¶</a></h2>
<p>All of deep learning is computations on tensors, which are
generalizations of a matrix that can be indexed in more than 2
dimensions. We will see exactly what this means in-depth later. First,
let’s look what we can do with tensors.</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="c1"># Author: Robert Guthrie</span>
<span class="kn">import</span> <span class="nn">torch</span>
<span class="n">torch</span><span class="o">.</span><span class="n">manual_seed</span><span class="p">(</span><span class="mi">1</span><span class="p">)</span>
</pre></div>
</div>
<div class="section" id="creating-tensors">
<h3>Creating Tensors<a class="headerlink" href="#creating-tensors" title="Permalink to this headline">¶</a></h3>
<p>Tensors can be created from Python lists with the torch.tensor()
function.</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="c1"># torch.tensor(data) creates a torch.Tensor object with the given data.</span>
<span class="n">V_data</span> <span class="o">=</span> <span class="p">[</span><span class="mf">1.</span><span class="p">,</span> <span class="mf">2.</span><span class="p">,</span> <span class="mf">3.</span><span class="p">]</span>
<span class="n">V</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">tensor</span><span class="p">(</span><span class="n">V_data</span><span class="p">)</span>
<span class="nb">print</span><span class="p">(</span><span class="n">V</span><span class="p">)</span>
<span class="c1"># Creates a matrix</span>
<span class="n">M_data</span> <span class="o">=</span> <span class="p">[[</span><span class="mf">1.</span><span class="p">,</span> <span class="mf">2.</span><span class="p">,</span> <span class="mf">3.</span><span class="p">],</span> <span class="p">[</span><span class="mf">4.</span><span class="p">,</span> <span class="mf">5.</span><span class="p">,</span> <span class="mi">6</span><span class="p">]]</span>
<span class="n">M</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">tensor</span><span class="p">(</span><span class="n">M_data</span><span class="p">)</span>
<span class="nb">print</span><span class="p">(</span><span class="n">M</span><span class="p">)</span>
<span class="c1"># Create a 3D tensor of size 2x2x2.</span>
<span class="n">T_data</span> <span class="o">=</span> <span class="p">[[[</span><span class="mf">1.</span><span class="p">,</span> <span class="mf">2.</span><span class="p">],</span> <span class="p">[</span><span class="mf">3.</span><span class="p">,</span> <span class="mf">4.</span><span class="p">]],</span>
<span class="p">[[</span><span class="mf">5.</span><span class="p">,</span> <span class="mf">6.</span><span class="p">],</span> <span class="p">[</span><span class="mf">7.</span><span class="p">,</span> <span class="mf">8.</span><span class="p">]]]</span>
<span class="n">T</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">tensor</span><span class="p">(</span><span class="n">T_data</span><span class="p">)</span>
<span class="nb">print</span><span class="p">(</span><span class="n">T</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>tensor([1., 2., 3.])
tensor([[1., 2., 3.],
[4., 5., 6.]])
tensor([[[1., 2.],
[3., 4.]],
[[5., 6.],
[7., 8.]]])
</pre></div>
</div>
<p>What is a 3D tensor anyway? Think about it like this. If you have a
vector, indexing into the vector gives you a scalar. If you have a
matrix, indexing into the matrix gives you a vector. If you have a 3D
tensor, then indexing into the tensor gives you a matrix!</p>
<p>A note on terminology:
when I say “tensor” in this tutorial, it refers
to any torch.Tensor object. Matrices and vectors are special cases of
torch.Tensors, where their dimension is 2 and 1 respectively. When I am
talking about 3D tensors, I will explicitly use the term “3D tensor”.</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="c1"># Index into V and get a scalar (0 dimensional tensor)</span>
<span class="nb">print</span><span class="p">(</span><span class="n">V</span><span class="p">[</span><span class="mi">0</span><span class="p">])</span>
<span class="c1"># Get a Python number from it</span>
<span class="nb">print</span><span class="p">(</span><span class="n">V</span><span class="p">[</span><span class="mi">0</span><span class="p">]</span><span class="o">.</span><span class="n">item</span><span class="p">())</span>
<span class="c1"># Index into M and get a vector</span>
<span class="nb">print</span><span class="p">(</span><span class="n">M</span><span class="p">[</span><span class="mi">0</span><span class="p">])</span>
<span class="c1"># Index into T and get a matrix</span>
<span class="nb">print</span><span class="p">(</span><span class="n">T</span><span class="p">[</span><span class="mi">0</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>tensor(1.)
1.0
tensor([1., 2., 3.])
tensor([[1., 2.],
[3., 4.]])
</pre></div>
</div>
<p>You can also create tensors of other data types. To create a tensor of integer types, try
torch.tensor([[1, 2], [3, 4]]) (where all elements in the list are integers).
You can also specify a data type by passing in <code class="docutils literal notranslate"><span class="pre">dtype=torch.data_type</span></code>.
Check the documentation for more data types, but
Float and Long will be the most common.</p>
<p>You can create a tensor with random data and the supplied dimensionality
with torch.randn()</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">x</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">((</span><span class="mi">3</span><span class="p">,</span> <span class="mi">4</span><span class="p">,</span> <span class="mi">5</span><span class="p">))</span>
<span class="nb">print</span><span class="p">(</span><span class="n">x</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>tensor([[[-1.5256, -0.7502, -0.6540, -1.6095, -0.1002],
[-0.6092, -0.9798, -1.6091, -0.7121, 0.3037],
[-0.7773, -0.2515, -0.2223, 1.6871, 0.2284],
[ 0.4676, -0.6970, -1.1608, 0.6995, 0.1991]],
[[ 0.8657, 0.2444, -0.6629, 0.8073, 1.1017],
[-0.1759, -2.2456, -1.4465, 0.0612, -0.6177],
[-0.7981, -0.1316, 1.8793, -0.0721, 0.1578],
[-0.7735, 0.1991, 0.0457, 0.1530, -0.4757]],
[[-0.1110, 0.2927, -0.1578, -0.0288, 0.4533],
[ 1.1422, 0.2486, -1.7754, -0.0255, -1.0233],
[-0.5962, -1.0055, 0.4285, 1.4761, -1.7869],
[ 1.6103, -0.7040, -0.1853, -0.9962, -0.8313]]])
</pre></div>
</div>
</div>
<div class="section" id="operations-with-tensors">
<h3>Operations with Tensors<a class="headerlink" href="#operations-with-tensors" title="Permalink to this headline">¶</a></h3>
<p>You can operate on tensors in the ways you would expect.</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">x</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">tensor</span><span class="p">([</span><span class="mf">1.</span><span class="p">,</span> <span class="mf">2.</span><span class="p">,</span> <span class="mf">3.</span><span class="p">])</span>
<span class="n">y</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">tensor</span><span class="p">([</span><span class="mf">4.</span><span class="p">,</span> <span class="mf">5.</span><span class="p">,</span> <span class="mf">6.</span><span class="p">])</span>
<span class="n">z</span> <span class="o">=</span> <span class="n">x</span> <span class="o">+</span> <span class="n">y</span>
<span class="nb">print</span><span class="p">(</span><span class="n">z</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>tensor([5., 7., 9.])
</pre></div>
</div>
<p>See <a class="reference external" href="https://pytorch.org/docs/torch.html">the documentation</a> for a
complete list of the massive number of operations available to you. They
expand beyond just mathematical operations.</p>
<p>One helpful operation that we will make use of later is concatenation.</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="c1"># By default, it concatenates along the first axis (concatenates rows)</span>
<span class="n">x_1</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span> <span class="mi">5</span><span class="p">)</span>
<span class="n">y_1</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">3</span><span class="p">,</span> <span class="mi">5</span><span class="p">)</span>
<span class="n">z_1</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">cat</span><span class="p">([</span><span class="n">x_1</span><span class="p">,</span> <span class="n">y_1</span><span class="p">])</span>
<span class="nb">print</span><span class="p">(</span><span class="n">z_1</span><span class="p">)</span>
<span class="c1"># Concatenate columns:</span>
<span class="n">x_2</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span> <span class="mi">3</span><span class="p">)</span>
<span class="n">y_2</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span> <span class="mi">5</span><span class="p">)</span>
<span class="c1"># second arg specifies which axis to concat along</span>
<span class="n">z_2</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">cat</span><span class="p">([</span><span class="n">x_2</span><span class="p">,</span> <span class="n">y_2</span><span class="p">],</span> <span class="mi">1</span><span class="p">)</span>
<span class="nb">print</span><span class="p">(</span><span class="n">z_2</span><span class="p">)</span>
<span class="c1"># If your tensors are not compatible, torch will complain. Uncomment to see the error</span>
<span class="c1"># torch.cat([x_1, x_2])</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>tensor([[-0.8029, 0.2366, 0.2857, 0.6898, -0.6331],
[ 0.8795, -0.6842, 0.4533, 0.2912, -0.8317],
[-0.5525, 0.6355, -0.3968, -0.6571, -1.6428],
[ 0.9803, -0.0421, -0.8206, 0.3133, -1.1352],
[ 0.3773, -0.2824, -2.5667, -1.4303, 0.5009]])
tensor([[ 0.5438, -0.4057, 1.1341, -0.1473, 0.6272, 1.0935, 0.0939, 1.2381],
[-1.1115, 0.3501, -0.7703, -1.3459, 0.5119, -0.6933, -0.1668, -0.9999]])
</pre></div>
</div>
</div>
<div class="section" id="reshaping-tensors">
<h3>Reshaping Tensors<a class="headerlink" href="#reshaping-tensors" title="Permalink to this headline">¶</a></h3>
<p>Use the .view() method to reshape a tensor. This method receives heavy
use, because many neural network components expect their inputs to have
a certain shape. Often you will need to reshape before passing your data
to the component.</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">x</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span> <span class="mi">3</span><span class="p">,</span> <span class="mi">4</span><span class="p">)</span>
<span class="nb">print</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
<span class="nb">print</span><span class="p">(</span><span class="n">x</span><span class="o">.</span><span class="n">view</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span> <span class="mi">12</span><span class="p">))</span> <span class="c1"># Reshape to 2 rows, 12 columns</span>
<span class="c1"># Same as above. If one of the dimensions is -1, its size can be inferred</span>
<span class="nb">print</span><span class="p">(</span><span class="n">x</span><span class="o">.</span><span class="n">view</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span> <span class="o">-</span><span class="mi">1</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>tensor([[[ 0.4175, -0.2127, -0.8400, -0.4200],
[-0.6240, -0.9773, 0.8748, 0.9873],
[-0.0594, -2.4919, 0.2423, 0.2883]],
[[-0.1095, 0.3126, 1.5038, 0.5038],
[ 0.6223, -0.4481, -0.2856, 0.3880],
[-1.1435, -0.6512, -0.1032, 0.6937]]])
tensor([[ 0.4175, -0.2127, -0.8400, -0.4200, -0.6240, -0.9773, 0.8748, 0.9873,
-0.0594, -2.4919, 0.2423, 0.2883],
[-0.1095, 0.3126, 1.5038, 0.5038, 0.6223, -0.4481, -0.2856, 0.3880,
-1.1435, -0.6512, -0.1032, 0.6937]])
tensor([[ 0.4175, -0.2127, -0.8400, -0.4200, -0.6240, -0.9773, 0.8748, 0.9873,
-0.0594, -2.4919, 0.2423, 0.2883],
[-0.1095, 0.3126, 1.5038, 0.5038, 0.6223, -0.4481, -0.2856, 0.3880,
-1.1435, -0.6512, -0.1032, 0.6937]])
</pre></div>
</div>
</div>
</div>
<div class="section" id="computation-graphs-and-automatic-differentiation">
<h2>Computation Graphs and Automatic Differentiation<a class="headerlink" href="#computation-graphs-and-automatic-differentiation" title="Permalink to this headline">¶</a></h2>
<p>The concept of a computation graph is essential to efficient deep
learning programming, because it allows you to not have to write the
back propagation gradients yourself. A computation graph is simply a
specification of how your data is combined to give you the output. Since
the graph totally specifies what parameters were involved with which
operations, it contains enough information to compute derivatives. This
probably sounds vague, so let’s see what is going on using the
fundamental flag <code class="docutils literal notranslate"><span class="pre">requires_grad</span></code>.</p>
<p>First, think from a programmers perspective. What is stored in the
torch.Tensor objects we were creating above? Obviously the data and the
shape, and maybe a few other things. But when we added two tensors
together, we got an output tensor. All this output tensor knows is its
data and shape. It has no idea that it was the sum of two other tensors
(it could have been read in from a file, it could be the result of some
other operation, etc.)</p>
<p>If <code class="docutils literal notranslate"><span class="pre">requires_grad=True</span></code>, the Tensor object keeps track of how it was
created. Let’s see it in action.</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="c1"># Tensor factory methods have a ``requires_grad`` flag</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">tensor</span><span class="p">([</span><span class="mf">1.</span><span class="p">,</span> <span class="mf">2.</span><span class="p">,</span> <span class="mi">3</span><span class="p">],</span> <span class="n">requires_grad</span><span class="o">=</span><span class="kc">True</span><span class="p">)</span>
<span class="c1"># With requires_grad=True, you can still do all the operations you previously</span>
<span class="c1"># could</span>
<span class="n">y</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">tensor</span><span class="p">([</span><span class="mf">4.</span><span class="p">,</span> <span class="mf">5.</span><span class="p">,</span> <span class="mi">6</span><span class="p">],</span> <span class="n">requires_grad</span><span class="o">=</span><span class="kc">True</span><span class="p">)</span>
<span class="n">z</span> <span class="o">=</span> <span class="n">x</span> <span class="o">+</span> <span class="n">y</span>
<span class="nb">print</span><span class="p">(</span><span class="n">z</span><span class="p">)</span>
<span class="c1"># BUT z knows something extra.</span>
<span class="nb">print</span><span class="p">(</span><span class="n">z</span><span class="o">.</span><span class="n">grad_fn</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>tensor([5., 7., 9.], grad_fn=<AddBackward0>)
<AddBackward0 object at 0x7f2891924250>
</pre></div>
</div>
<p>So Tensors know what created them. z knows that it wasn’t read in from
a file, it wasn’t the result of a multiplication or exponential or
whatever. And if you keep following z.grad_fn, you will find yourself at
x and y.</p>
<p>But how does that help us compute a gradient?</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="c1"># Let's sum up all the entries in z</span>
<span class="n">s</span> <span class="o">=</span> <span class="n">z</span><span class="o">.</span><span class="n">sum</span><span class="p">()</span>
<span class="nb">print</span><span class="p">(</span><span class="n">s</span><span class="p">)</span>
<span class="nb">print</span><span class="p">(</span><span class="n">s</span><span class="o">.</span><span class="n">grad_fn</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>tensor(21., grad_fn=<SumBackward0>)
<SumBackward0 object at 0x7f2891924c70>
</pre></div>
</div>
<p>So now, what is the derivative of this sum with respect to the first
component of x? In math, we want</p>
<div class="math">
\[\frac{\partial s}{\partial x_0}\]</div>
<p>Well, s knows that it was created as a sum of the tensor z. z knows
that it was the sum x + y. So</p>
<div class="math">
\[s = \overbrace{x_0 + y_0}^\text{$z_0$} + \overbrace{x_1 + y_1}^\text{$z_1$} + \overbrace{x_2 + y_2}^\text{$z_2$}
\]</div>
<p>And so s contains enough information to determine that the derivative
we want is 1!</p>
<p>Of course this glosses over the challenge of how to actually compute
that derivative. The point here is that s is carrying along enough
information that it is possible to compute it. In reality, the
developers of Pytorch program the sum() and + operations to know how to
compute their gradients, and run the back propagation algorithm. An
in-depth discussion of that algorithm is beyond the scope of this
tutorial.</p>
<p>Let’s have Pytorch compute the gradient, and see that we were right:
(note if you run this block multiple times, the gradient will increment.
That is because Pytorch <em>accumulates</em> the gradient into the .grad
property, since for many models this is very convenient.)</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="c1"># calling .backward() on any variable will run backprop, starting from it.</span>
<span class="n">s</span><span class="o">.</span><span class="n">backward</span><span class="p">()</span>
<span class="nb">print</span><span class="p">(</span><span class="n">x</span><span class="o">.</span><span class="n">grad</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>tensor([1., 1., 1.])
</pre></div>
</div>
<p>Understanding what is going on in the block below is crucial for being a
successful programmer in deep learning.</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">x</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span> <span class="mi">2</span><span class="p">)</span>
<span class="n">y</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span> <span class="mi">2</span><span class="p">)</span>
<span class="c1"># By default, user created Tensors have ``requires_grad=False``</span>
<span class="nb">print</span><span class="p">(</span><span class="n">x</span><span class="o">.</span><span class="n">requires_grad</span><span class="p">,</span> <span class="n">y</span><span class="o">.</span><span class="n">requires_grad</span><span class="p">)</span>
<span class="n">z</span> <span class="o">=</span> <span class="n">x</span> <span class="o">+</span> <span class="n">y</span>
<span class="c1"># So you can't backprop through z</span>
<span class="nb">print</span><span class="p">(</span><span class="n">z</span><span class="o">.</span><span class="n">grad_fn</span><span class="p">)</span>
<span class="c1"># ``.requires_grad_( ... )`` changes an existing Tensor's ``requires_grad``</span>
<span class="c1"># flag in-place. The input flag defaults to ``True`` if not given.</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">x</span><span class="o">.</span><span class="n">requires_grad_</span><span class="p">()</span>
<span class="n">y</span> <span class="o">=</span> <span class="n">y</span><span class="o">.</span><span class="n">requires_grad_</span><span class="p">()</span>
<span class="c1"># z contains enough information to compute gradients, as we saw above</span>
<span class="n">z</span> <span class="o">=</span> <span class="n">x</span> <span class="o">+</span> <span class="n">y</span>
<span class="nb">print</span><span class="p">(</span><span class="n">z</span><span class="o">.</span><span class="n">grad_fn</span><span class="p">)</span>
<span class="c1"># If any input to an operation has ``requires_grad=True``, so will the output</span>
<span class="nb">print</span><span class="p">(</span><span class="n">z</span><span class="o">.</span><span class="n">requires_grad</span><span class="p">)</span>
<span class="c1"># Now z has the computation history that relates itself to x and y</span>
<span class="c1"># Can we just take its values, and **detach** it from its history?</span>
<span class="n">new_z</span> <span class="o">=</span> <span class="n">z</span><span class="o">.</span><span class="n">detach</span><span class="p">()</span>
<span class="c1"># ... does new_z have information to backprop to x and y?</span>
<span class="c1"># NO!</span>
<span class="nb">print</span><span class="p">(</span><span class="n">new_z</span><span class="o">.</span><span class="n">grad_fn</span><span class="p">)</span>
<span class="c1"># And how could it? ``z.detach()`` returns a tensor that shares the same storage</span>
<span class="c1"># as ``z``, but with the computation history forgotten. It doesn't know anything</span>
<span class="c1"># about how it was computed.</span>
<span class="c1"># In essence, we have broken the Tensor away from its past history</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>False False
None
<AddBackward0 object at 0x7f2891db0b20>
True
None
</pre></div>
</div>
<p>You can also stop autograd from tracking history on Tensors
with <code class="docutils literal notranslate"><span class="pre">.requires_grad=True</span></code> by wrapping the code block in
<code class="docutils literal notranslate"><span class="pre">with</span> <span class="pre">torch.no_grad():</span></code></p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="nb">print</span><span class="p">(</span><span class="n">x</span><span class="o">.</span><span class="n">requires_grad</span><span class="p">)</span>
<span class="nb">print</span><span class="p">((</span><span class="n">x</span> <span class="o">**</span> <span class="mi">2</span><span class="p">)</span><span class="o">.</span><span class="n">requires_grad</span><span class="p">)</span>
<span class="k">with</span> <span class="n">torch</span><span class="o">.</span><span class="n">no_grad</span><span class="p">():</span>
<span class="nb">print</span><span class="p">((</span><span class="n">x</span> <span class="o">**</span> <span class="mi">2</span><span class="p">)</span><span class="o">.</span><span class="n">requires_grad</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>True
True
False
</pre></div>
</div>
<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 0 minutes 0.144 seconds)</p>
<div class="sphx-glr-footer class sphx-glr-footer-example docutils container" id="sphx-glr-download-beginner-nlp-pytorch-tutorial-py">
<div class="sphx-glr-download docutils container">
<a class="reference download internal" download="" href="../../_downloads/dc8f8f6db5300c8bbf5454f3e8c0dfcd/pytorch_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">pytorch_tutorial.py</span></code></a></div>
<div class="sphx-glr-download docutils container">
<a class="reference download internal" download="" href="../../_downloads/526427246f9f04fd3eb099935090d87c/pytorch_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">pytorch_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>
</article>
</div>
<footer>
<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>
© Copyright 2021, PyTorch & 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="#">Introduction to PyTorch</a><ul>
<li><a class="reference internal" href="#introduction-to-torch-s-tensor-library">Introduction to Torch’s tensor library</a><ul>
<li><a class="reference internal" href="#creating-tensors">Creating Tensors</a></li>
<li><a class="reference internal" href="#operations-with-tensors">Operations with Tensors</a></li>
<li><a class="reference internal" href="#reshaping-tensors">Reshaping Tensors</a></li>
</ul>
</li>
<li><a class="reference internal" href="#computation-graphs-and-automatic-differentiation">Computation Graphs and Automatic Differentiation</a></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>