d2l-jax
JAX/Flax implementation for Dive into Deep Learning.
- 1. Introduction
- 2. Preliminaries
- 3. Linear Neural Networks for Regression
- 4. Linear Neural Networks for Classification
- 5. Multilayer Perceptrons
- 6. Builders’ Guide
- 7. Convolutional Neural Networks
- 8. Modern Convolutional Neural Networks (unfinished)
- [WIP] 9. Recurrent Neural Networks
- 10. Modern Recurrent Neural Networks
- 11. Attention Mechanisms and Transformers
- 12. Optimization Algorithms
- 13. Computational Performance
- 14. Computer Vision
- 15. Natural Language Processing: Pretraining
- 16. Natural Language Processing: Applications
- 17. Recommender Systems
- 18. Generative Adversarial Networks
- 19. Appendix: Mathematics for Deep Learning
- 20. Appendix: Tools for Deep Learning
Todos
- apply vmap & jit
- Set Dropout determinism based on on a variable