Skip to content

Neulus/trl

Repository files navigation

Some hacky fork of TRL.

Forked to enable Whisper GRPO training for various purposes. Not suitable for production use.

Code probably won't work beside Whisper models. Also probably won't work for whisper-large-v3 (Due to new sampling_rate).

About

Train transformer language & seq2seq(Whisper only) models with reinforcement learning.

Resources

License

Code of conduct

Contributing

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages