All Projects → wzy6642 → Dive Into Deep Learning Pytorch Pdf

wzy6642 / Dive Into Deep Learning Pytorch Pdf

Licence: gpl-3.0
本项目对中文版《动手学深度学习》中的代码进行了PyTorch实现并整理为PDF版本供下载

Projects that are alternatives of or similar to Dive Into Deep Learning Pytorch Pdf

Machinelearningnote
Stars: ✭ 142 (-1.39%)
Mutual labels:  jupyter-notebook
Dlfs code
Code for the book Deep Learning From Scratch, from O'Reilly September 2019
Stars: ✭ 142 (-1.39%)
Mutual labels:  jupyter-notebook
Practicalai Cn
AI实战-practicalAI 中文版
Stars: ✭ 2,375 (+1549.31%)
Mutual labels:  jupyter-notebook
Vmls Companions
These are companion notebooks written in Julia and Python for: "Introduction to Applied Linear Algebra" by Boyd and Vandenberghe.
Stars: ✭ 142 (-1.39%)
Mutual labels:  jupyter-notebook
Faster Rcnn tensorflow
This is a tensorflow re-implementation of Faster R-CNN: Towards Real-Time ObjectDetection with Region Proposal Networks.
Stars: ✭ 142 (-1.39%)
Mutual labels:  jupyter-notebook
Machine learning for good
Machine learning fundamentals lesson in interactive notebooks
Stars: ✭ 142 (-1.39%)
Mutual labels:  jupyter-notebook
Celeste.jl
Scalable inference for a generative model of astronomical images
Stars: ✭ 142 (-1.39%)
Mutual labels:  jupyter-notebook
Main
CS579: Online Social Network Analysis at the Illinois Institute of Technology
Stars: ✭ 143 (-0.69%)
Mutual labels:  jupyter-notebook
Selfconsistency
Code for the paper: Fighting Fake News: Image Splice Detection via Learned Self-Consistency
Stars: ✭ 143 (-0.69%)
Mutual labels:  jupyter-notebook
Visualizing cnns
Using Keras and cats to visualize layers from CNNs
Stars: ✭ 143 (-0.69%)
Mutual labels:  jupyter-notebook
Deep and machine learning projects
This Repository contains the list of various Machine and Deep Learning related projects. Related code and data files are available inside this folder. One can go through these projects to implement them in real life for specific use cases.
Stars: ✭ 141 (-2.08%)
Mutual labels:  jupyter-notebook
Design Of Experiment Python
Design-of-experiment (DOE) generator for science, engineering, and statistics
Stars: ✭ 143 (-0.69%)
Mutual labels:  jupyter-notebook
Part2
Stars: ✭ 143 (-0.69%)
Mutual labels:  jupyter-notebook
Disprcnn
Code release for Stereo 3D Object Detection via Shape Prior Guided Instance Disparity Estimation (CVPR 2020)
Stars: ✭ 142 (-1.39%)
Mutual labels:  jupyter-notebook
Data Science Question Answer
A repo for data science related questions and answers
Stars: ✭ 2,000 (+1288.89%)
Mutual labels:  jupyter-notebook
Invisible Cloak
Stars: ✭ 143 (-0.69%)
Mutual labels:  jupyter-notebook
Animl
Reproduction of "Model-Agnostic Meta-Learning" (MAML) and "Reptile".
Stars: ✭ 143 (-0.69%)
Mutual labels:  jupyter-notebook
Gp
A tutorial about Gaussian process regression
Stars: ✭ 141 (-2.08%)
Mutual labels:  jupyter-notebook
Complete Python Bootcamp
Lectures for Udemy - Complete Python Bootcamp Course
Stars: ✭ 1,879 (+1204.86%)
Mutual labels:  jupyter-notebook
Tutorial Softweightsharingfornncompression
A tutorial on 'Soft weight-sharing for Neural Network compression' published at ICLR2017
Stars: ✭ 143 (-0.69%)
Mutual labels:  jupyter-notebook

Dive-Into-Deep-Learning-PyTorch-PDF

封面

简介

  本项目对中文版《动手学深度学习》中的代码进行整理,并参考一些优秀的GitHub项目给出基于PyTorch的实现方法。为了方便阅读,本项目给出全书PyTorch版的PDF版本。欢迎大家Download,Star,Fork。除了原书内容外,我们还为每一章增加了本章附录,用于对该章节中用到的函数以及数学计算加以说明,除此之外还增加了语义分割网络(U-Net)的实现。书籍百度云链接:https://pan.baidu.com/s/1l8yDHVcB0FXPLH1nL542xA 密码:euqd
  原书作者:阿斯顿·张、李沐、扎卡里 C. 立顿、亚历山大 J. 斯莫拉以及其他社区贡献者。
  备注: d2lzh.py与其它代码需要放到同一个文件夹下。

目录

  • 1. 预备知识
    • 1.1 数据操作
    • 1.2 自动求梯度
    • 1.3 查阅文档
    • 1.4 本章附录
  • 2. 深度学习基础
    • 2.1 线性回归
    • 2.2 线性回归的从零开始实现
    • 2.3 线性回归的简洁实现
    • 2.4 softmax回归
    • 2.5 图像分类数据集(Fashion-MNIST)
    • 2.6 softmax回归的从零开始实现
    • 2.7 softmax回归的简洁实现
    • 2.8 多层感知机
    • 2.9 多层感知机的从零开始实现
    • 2.10 多层感知机的简洁实现
    • 2.11 模型选择、欠拟合和过拟合
    • 2.12 权重衰减
    • 2.13 丢弃法
    • 2.14 正向传播、反向传播和计算图
    • 2.15 数值稳定性和模型初始化
    • 2.16 实战Kaggle比赛:房价预测
    • 2.17 本章附录
  • 3. 深度学习计算
    • 3.1 模型构造
    • 3.2 模型参数的访问、初始化和共享
    • 3.3 自定义层
    • 3.4 读取和存储
    • 3.5 GPU计算
    • 3.6 本章附录
  • 4. 卷积神经网络
    • 4.1 二维卷积层
    • 4.2 填充和步幅
    • 4.3 多输入通道和多输出通道
    • 4.4 池化层
    • 4.5 卷积神经网络(LeNet)
    • 4.6 深度卷积神经网络(AlexNet)
    • 4.7 使用重复元素的网络(VGG)
    • 4.8 网络中的网络(NiN)
    • 4.9 含并行连结的网络(GoogLeNet)
    • 4.10 批量归一化
    • 4.11 残差网络(ResNet)
    • 4.12 稠密连接网络(DenseNet)
    • 4.13 本章附录
  • 5. 循环神经网络
    • 5.1 语言模型
    • 5.2 循环神经网络
    • 5.3 语言模型数据集(周杰伦专辑歌词)
    • 5.4 循环神经网络的从零开始实现
    • 5.5 循环神经网络的简洁实现
    • 5.6 通过时间反向传播
    • 5.7 门控循环单元(GRU)
    • 5.8 长短期记忆(LSTM)
    • 5.9 深度循环神经网络
    • 5.10 双向循环神经网络
    • 5.11 本章附录
  • 6. 优化算法
    • 6.1 优化与深度学习
    • 6.2 梯度下降和随机梯度下降
    • 6.3 小批量随机梯度下降
    • 6.4 动量法
    • 6.5 AdaGrad算法
    • 6.6 RMSProp算法
    • 6.7 AdaDelta算法
    • 6.8 Adam算法
    • 6.9 本章附录
  • 7. 计算性能
    • 7.1 命令式和符号式混合编程
    • 7.2 自动并行计算
    • 7.3 多GPU计算
    • 7.4 本章附录
  • 8. 计算机视觉
    • 8.1 图像增广
    • 8.2 微调
    • 8.3 目标检测和边界框
    • 8.4 锚框
    • 8.5 多尺度目标检测
    • 8.6 目标检测数据集(皮卡丘)
    • 8.7 单发多框检测(SSD)
    • 8.8 区域卷积神经网络(R-CNN)系列
    • 8.9 语义分割和数据集
    • 8.10 全卷积网络(FCN)
    • 8.11 样式迁移
    • 8.12 实战Kaggle比赛:图像分类(CIFAR-10)
    • 8.13 实战Kaggle比赛:狗的品种识别(ImageNet Dogs)
    • 8.14 语义分割网络(U-Net)
    • 8.15 本章附录
  • 9. 自然语言处理
    • 9.1 词嵌入(word2vec)
    • 9.2 近似训练
    • 9.3 word2vec的实现
    • 9.4 子词嵌入(fastText)
    • 9.5 全局向量的词嵌入(GloVe)
    • 9.6 求近义词和类比词
    • 9.7 文本情感分类:使用循环神经网络
    • 9.8 文本情感分类:使用卷积神经网络(textCNN)
    • 9.9 编码器—解码器(seq2seq)
    • 9.10 束搜索
    • 9.11 注意力机制
    • 9.12 机器翻译
    • 9.13 本章附录

环境

matplotlib==3.3.2
torch==1.1.0
torchvision==0.3.0
torchtext==0.4.0
CUDA Version==11.0

参考

本书PyTorch实现:Dive-into-DL-PyTorch
本书TendorFlow2.0实现:Dive-into-DL-TensorFlow2.0

原书地址

中文版:动手学深度学习 | Github仓库
English Version: Dive into Deep Learning | Github Repo

引用

如果您在研究中使用了这个项目请引用原书:

@book{zhang2019dive,
    title={Dive into Deep Learning},
    author={Aston Zhang and Zachary C. Lipton and Mu Li and Alexander J. Smola},
    note={\url{http://www.d2l.ai}},
    year={2020}
}
Note that the project description data, including the texts, logos, images, and/or trademarks, for each open source project belongs to its rightful owner. If you wish to add or remove any projects, please contact us at [email protected].