Jerzha/d2l-keras

d2l-ai tensorflow keras implementation (updating chapter 10 ...)

《Dive into Deep Learning》 Keras Implementation

• Origin Edition Book : www.d2l.ai
• Chinese Edition : zh.d2l.ai
• English MXNet Code： https://github.com/d2l-ai/d2l-en
• Chinese MXNet Code： https://github.com/d2l-ai/d2l-zh
• 中文说明和目录

Notes:

1. Dynamic computation graph is not supported in Keras, so I use numpy or tensorflow2.0 instead.
2. I didn't copy the book here. There's only titles and code here.
3. It use livelossplot for tracking training loss and acc.
4. Suggest opening chapter from contents as below, it use nbviewer to render jupyter.

Contents:

Chapter2 Preliminaries

• 2.2 Data Manipulation
• 2.3 Automatic Differentiation

Chapter3 Linear Neural Networks

• 3.2 Linear Regression
• 3.5 Fashion Mnist
• 3.6 Softmax Regression
• 3.9 Multilayer Perceptrons
• 3.11 Underfitting Overfitting
• 3.12 Weight Decay
• 3.13 Dropout
• 3.16 Kaggle House Prices

Chapter4 Deep Learning Computation

• 4.1 Layers And Blocks
• 4.2 Parameter Management
• 4.3 Deferred Initialization
• 4.4 Custom Layers
• 4.5 File IO
• 4.6 GPUs

Chapter5 Convolutional Neural Networks

• 5.5 Lenet
• 5.6 Alexnet
• 5.7 VGG
• 5.8 Network In Network
• 5.9 GoogleNet
• 5.11 Resnet
• 5.12 Densenet

Chapter6 Recurrent Neural Networks

• 6.3 Language Model Dataset
• 6.4 RNN
• 6.7 GRU
• 6.8 LSTM

Chapter7 Optimization Algorithms

• 7.1 Optimization And Deeplearning
• 7.2 Gradient Descent
• 7.4 Momentum
• 7.5 Adagrad
• 7.6 Rmsprop

Chapter9 Computer Vision

• 9.1 Image Augmentation
• 9.2 Fine Tuning
• 9.3 Bounding Box
• 9.4 Anchor Box
• 9.5 Multiscale Object Detection
• 9.6 Object Detection Dataset
• 9.7 SSD
• 9.8 RCNN
• 9.9 Semantic Segmentation
• 9.10 Fully Convolutional Network
• 9.11 Neural Style Transfer
• 9.12 Kaggle Cifar10

Chapter10 Natural Language Processing

• 10.3 Word2vec
• UPDATING...

Cite

BibTeX entry:

@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={2019}
}