Dive into Deep Learning: Sources and Full Index
Source and license
This directory was organized from D2L Chinese-site pages into Obsidian Markdown. Original site: D2L Chinese GitHub project: Dive into Deep Learning License: Apache-2.0
Local navigation status
On 2026-07-22, all 140 course entries were rechecked against the current nested file tree, with folder-qualified wikilinks. The D2L hierarchy and license provenance remain intact.
00-manifest.jsonno longer contains a historical absolute local path. A small number of high-risk entry points have explicitly labeled original 2026 engineering supplements; they are not presented as D2L source text.
English publication boundary
The links below point to license-safe English source-reference stubs, not translated republications of the frozen D2L bodies. Use the official source for the full licensed course text. For this knowledge base’s original English learning path and engineering guidance, see Deep Learning and Engineering Practice and a Modern Workflow.
Full index
- 01 Preface
- 02 Installation
- 03 Notation
- 04 Introduction
- 05 Preliminaries
- 06 Linear Neural Networks for Regression
- 07 Multilayer Perceptrons
- 04.1 Multilayer Perceptrons
- 04.2 Multilayer Perceptrons from Scratch
- 04.3 Concise Implementation of Multilayer Perceptrons
- 04.4 Model Selection, Underfitting, and Overfitting
- 04.5 Weight Decay
- 04.6 Dropout
- 04.7 Forward Propagation, Backpropagation, and Computational Graphs
- 04.8 Numerical Stability and Initialization
- 04.9 Environment and Distribution Shift
- 04.10 Kaggle House Price Prediction
- 08 Deep Learning Computation
- 09 Convolutional Neural Networks
- 10 Modern Convolutional Neural Networks
- 11 Recurrent Neural Networks
- 12 Modern Recurrent Neural Networks
- 13 Attention Mechanisms
- 14 Optimization Algorithms
- 15 Computational Performance
- 16 Computer Vision
- 13.1 Image Augmentation
- 13.2 Fine-Tuning
- 13.3 Object Detection and Bounding Boxes
- 13.4 Anchor Boxes
- 13.5 Multiscale Object Detection
- 13.6 Object Detection Dataset
- 13.7 Single Shot Multibox Detection (SSD)
- 13.8 Region-Based Convolutional Neural Network (R-CNN) Family
- 13.9 Semantic Segmentation and Datasets
- 13.10 Transposed Convolution
- 13.11 Fully Convolutional Networks
- 13.12 Neural Style Transfer
- 13.13 Kaggle Image Classification (CIFAR-10)
- 13.14 Kaggle Dog Breed Identification (ImageNet Dogs)
- 17 Natural Language Processing: Pretraining
- 14.1 Word Embedding (word2vec)
- 14.2 Approximate Training
- 14.3 Dataset for Pretraining Word Embeddings
- 14.4 Pretraining word2vec
- 14.5 Global Vectors for Word Representation (GloVe)
- 14.6 Subword Embedding
- 14.7 Word Similarity and Analogy Tasks
- 14.8 Bidirectional Encoder Representations from Transformers (BERT)
- 14.9 Dataset for Pretraining BERT
- 14.10 Pretraining BERT
- 18 Natural Language Processing: Applications
- 15.1 Sentiment Analysis and Datasets
- 15.2 Sentiment Analysis with Recurrent Neural Networks
- 15.3 Sentiment Analysis with Convolutional Neural Networks
- 15.4 Natural Language Inference and Datasets
- 15.5 Natural Language Inference with Attention
- 15.6 Fine-Tuning BERT for Sequence-Level and Token-Level Applications
- 15.7 Natural Language Inference: Fine-Tuning BERT
- 19 Appendix: Deep Learning Tools
- 20 References