This is the support site for the book “Deep Learning from Scratch 2: Natural Language Processing Edition” (O'Reilly Japan). The source code used in the book is collected here.
| Folder | Description |
|---|---|
| ch01 | Source code for Chapter 1 |
| ch02 | Source code for Chapter 2 |
| ... | ... |
| ch08 | Source code for Chapter 8 |
| common | Common source code |
| dataset | Source code for datasets |
Pre-trained weight files (used in Chapters 6 and 7) can be downloaded from the following URL:
https://www.oreilly.co.jp/pub/9784873118369/BetterRnnlm.pkl
For explanations of the source code, please refer to the book.
To run the source code, you need the following software:
- Python 3.x (version 3 series)
- NumPy
- Matplotlib
Optionally, you can also use the following libraries:
- SciPy (optional)
- CuPy (optional)
Move to the folder of each chapter and run the Python command.
$ cd ch01
$ python train.py
$ cd ../ch05
$ python train_custom_loop.py
The source code in this repository is licensed under the MIT License.
You are free to use it for both commercial and non-commercial purposes.
Errata for the book are published at the following page:
https://github.com/oreilly-japan/deep-learning-from-scratch-2/wiki/errata
If you find any errors not listed on this page, please contact japan@oreilly.co.jp.