Thank you for looking at contributing to this repository! Anybody and everybody is welcome to do so :)
If you've found an issue with the code, and an issue hasn't been opened concerning it already, create an issue yourself!
If you can fix an issue with the code, open a Pull Request.
Use Discussions if you:
- Have any questions (
Q&Atab) - Have any ideas, mainly for what to do to improve the tokenization layer (
Ideastab) - Have something generally to say (
Generaltab)
Make a Pull Request for fixing an Issue, or adding something new, such as making new experiments (like train_imdb.py),
adding new results (such as checkpoints of trained models), or even changing the core TokenizationLayer itself.