pytorch handbook

PyTorch Chinese Manual (pytorch handbook)

Book introduction

This is an open source book with the goal of helping those who wish to and use PyTorch for deep learning development and research quickly.

Due to my limited level, I have referred to some online information when writing this tutorial. I would like to express my respect to them here. I will attach the original address to each quotation for your reference.

Deep learning technology is developing rapidly, and PyTorch is also constantly updated, and I will gradually improve the relevant content.

Version Description

Due to the change of PyTorch version, the version of the tutorial will be consistent with the PyTorch version.

Summary of the main changes in the pytorch major version update current version 1.11

Online version and PDF

Domestic mirrors are very fast and will not be blocked: https://www.pytorch.wiki/

There is no good method to generate PDF files yet. Friends who are familiar with this aspect can contact me. I am grateful.

QQ 6 Groups

Group number: 760443051

Click the link to join the group chat [PyTorch Handbook Communication Group 6]: https://jq.qq.com/?_wv=1027&k=X4Ro6uWv

Group 1 (985896536) is full, group 2 (681980831) Group 3 (773681699) is full 4 (884017356) is full 5 (894059877) is full

Don't add it

New Benefits

Public account daily sharing of dry goods articles

illustrate

• Please directly mention the issue or PR when modifying the typo

• Please pay attention to the version when PR

• If you have any questions, please directly ask the issue

grateful

Table of contents

Chapter 1: Getting Started with PyTorch

1. Introduction to PyTorch
2. PyTorch environment construction
3. PyTorch Deep Learning: 60-minute Quick Start (Official)
   • Tensor
   • Autograd: Automatically search for guidance
   • Neural Network
   • Training a classifier
   • Selected reading: parallel data processing (multiple GPU)
4. Related resources introduction

Chapter 2 Basics

Section 1 PyTorch Basics

1. Tensor
2. Automatic guide
3. Neural network package nn and optimizer optm
4. Data loading and preprocessing

Section 2 Basics of Deep Learning and Mathematical Principles

The basics of deep learning and mathematical principles

Section 3 Introduction to neural networks

Introduction to Neural Network Note: This chapter will crash when opening using Microsoft's Edge locally. Please enable Chrome Firefox to open to view.

Section 4 Convolutional Neural Network

Convolutional neural network

Section 5 Recurrent Neural Network

Recurrent neural network

Chapter 3 Practice

Section 1 logistic regression binary classification

logistic regression binary classification

Section 2 CNN: Handwritten numerical recognition of MNIST dataset

CNN: Handwritten numerical recognition of MNIST dataset

Section 3 RNN Example: Prediction of Cos by Sin

RNN instance: Prediction of Cos by Sin

Chapter 4 Improvement

Section 1 Fine-tuning

Fine-tuning

Section 2 Visualization

visdom

tensorboardx

Visually understand convolutional neural networks

Section 3 Fast.ai

Fast.ai

Some tips for training in Section 4

Section 5 Multi-GPU parallel training

Multi-GPU parallel computing

Supplementary translation article: Using DistributedDataParallel in PyTorch for multi-GPU distributed model training

Use DistributedDataParallel in PyTorch for Multi-GPU distributed model training

Chapter 5 Application

Section 1 Kaggle introduction

Introduction to Kaggle

Section 2 Structured Data

Pytorch handles structured data

Section 3 Computer Vision

Fashion MNIST Image Classification

Section 4 Natural Language Processing

Section 5 Collaborative Filtering

Chapter 6 Resources

torchaudio

Chapter 7 Appendix

Raspberry Pi compilation and installation of pytorch 1.4

Summary of common operations of transforms

Summary of loss function of pytorch

pytorch optimizer summary

Script

The script directory is a script I wrote to convert ipynb into online version and pdf file. Because it is still in the testing stage, please ask if you have any questions.

License

This work is licensed under the Creative Commons Attribution-Non-Commercial-Share-Share 3.0 Mainland China License Agreement