"Deep learning framework PyTorch: Getting Started and practice _ Chen Yun (a)."
Baidu network disk link: https://pan.baidu.com/s/1XwzebG5biDgC3ZY6g9nyvA extraction code: 2kac
brief introduction · · · · · ·
"Deep learning framework PyTorch: Getting Started and Practice" From multi-dimensional array Tensor, takes the reader step by step to understand the basics of all aspects of PyTorch. Combined with the basic knowledge and cutting-edge research, takes the reader from scratch to complete several classic interesting depth study of small projects, including the GAN generated animation picture, AI filter, AI poetry and so on. "Deep learning framework PyTorch: Getting Started and practice" is not a simple mechanical introduction to the use of each function interface, but try categorized, step by step introduction to the knowledge PyTorch the reader, I hope readers PyTorch have a complete understanding.
"Deep learning framework PyTorch: Getting Started and practice" content Deep, deep learning whether beginner or first contact with researchers PyTorch, can quickly grasp PyTorch in the process of learning this book. Even if there is some PyTorch user experience, but also be able to obtain PyTorch not the same as understanding from the book.
About the Author · · · · · ·
Chen Yun
Python programmers, Linux enthusiasts and PyTorch source contributors. The main research interests include computer vision and machine learning. "I know almost see the mountains, 2017 Machine Learning Challenge Cup" award, "2017 Tianchi medical AI Competition" eighth. Keen to promote PyTorch, and has a wealth of experience, active in the forum and know almost PyTorch related sectors.
table of Contents......
1 PyTorch简介
1.1 PyTorch的诞生
1.2 常见的深度学习框架简介
1.2.1 Theano
1.2.2 TensorFlow
1.2.3 Keras
1.2.4 Caffe/Caffe2
1.2.5 MXNet
1.2.6 CNTK
1.2.7 其他框架
1.3 属于动态图的未来
1.4 为什么选择PyTorch
1.5 星火燎原
1.6 fast.ai 放弃Keras+TensorFlow选择PyTorch
2 快速入门
2.1 安装与配置
2.1.1 安装PyTorch
2.1.2 学习环境配置
2.2 PyTorch入门第一步
2.2.1 Tensor
2.2.2 Autograd：自动微分
2.2.3 神经网络
2.2.4 小试牛刀：CIFAR-10分类
3 Tensor和autograd
3.1 Tensor
3.1.1 基础操作
3.1.2 Tensor和Numpy
3.1.3 内部结构
3.1.4 其他有关Tensor的话题
3.1.5 小试牛刀：线性回归
3.2 autograd
3.2.1 Variable
3.2.2 计算图
3.2.3 扩展autograd
3.2.4 小试牛刀：用Variable实现线性回归
4 神经网络工具箱nn
4.1 nn.Module
4.2 常用的神经网络层
4.2.1 图像相关层
4.2.2 激活函数
4.2.3 循环神经网络层
4.2.4 损失函数
4.3 优化器
4.4 nn.functional
4.5 初始化策略
4.6 nn.Module深入分析
4.7 nn和autograd的关系
4.8 小试牛刀：用50行代码搭建ResNet
5 PyTorch中常用的工具
5.1 数据处理
5.2 计算机视觉工具包：torchvision
5.3 可视化工具
5.3.1 Tensorboard
5.3.2 visdom
5.4 使用GPU加速：cuda
5.5 持久化
6 PyTorch实战指南
6.1 编程实战：猫和狗二分类
6.1.1 比赛介绍
6.1.2 文件组织架构
6.1.3 关于__init__.py
6.1.4 数据加载
6.1.5 模型定义
6.1.6 工具函数
6.1.7 配置文件
6.1.8 main.py
6.1.9 使用
6.1.10 争议
6.2 PyTorch Debug 指南
6.2.1 ipdb 介绍
6.2.2 在PyTorch中Debug
7 AI插画师：生成对抗网络
7.1 GAN的原理简介
7.2 用GAN生成动漫头像
7.3 实验结果分析
8 AI艺术家：神经网络风格迁移
8.1 风格迁移原理介绍
8.2 用PyTorch实现风格迁移
8.3 实验结果分析
9 AI诗人：用RNN写诗
9.1 自然语言处理的基础知识
9.1.1 词向量
9.1.2 RNN
9.2 CharRNN
9.3 用PyTorch实现CharRNN
9.4 实验结果分析
10 Image Caption：让神经网络看图讲故事
10.1 图像描述介绍
10.2 数据
10.2.1 数据介绍
10.2.2 图像数据处理
10.2.3 数据加载
10.3 模型与训练
10.4 实验结果分析
11 展望与未来
11.1 PyTorch的局限与发展
11.2 使用建议