lvwerra/trl

Train transformer language models with reinforcement learning.

3 best points · 0 comments · 2 Hacker News posts · 19,675 GitHub stars · View on GitHub

Hacker News discussions

• Transformer Reinforcement Learning (3 points, 0 comments, by fzliu, 2023-01-11)
• trl: Train transformer language models with reinforcement learning (1 points, 0 comments, by tosh, 2022-12-07)