Fix Gemma-4 GRPO catastrophic KL divergence with TRL 1.0.0+#21
Closed
danielhanchen wants to merge 2 commits into
Closed
Fix Gemma-4 GRPO catastrophic KL divergence with TRL 1.0.0+#21danielhanchen wants to merge 2 commits into
danielhanchen wants to merge 2 commits into
![pre-commit-ci[bot]](https://avatars.githubusercontent.com/in/68672?v=4&size=32){kind=avatar}