Skip to content

Do not force set beta to 0 for DAPO #3604

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Nov 17, 2025
Merged

Do not force set beta to 0 for DAPO #3604

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Do not force set beta to 0 for DAPO
  • Loading branch information
@Datta0
Datta0 committed Nov 17, 2025
commit 0e6e8e7f4c9fa1bb8c6e484414aad215a38ae93f
3 changes: 1 addition & 2 deletions unsloth/models/rl.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -814,10 +814,9 @@ def _patch_trl_rl_trainers(trainer_file = "grpo_trainer"):
" if epsilon_high != 0.28:\n"
" print('Unsloth: The DAPO paper recommends `epsilon_high = 0.28` - we will set it.')\n"
" if beta != 0.0:\n"
" print('Unsloth: The DAPO paper recommends setting `beta = 0.0` to remove the KL term - we will set it.')\n"
" print(f'[WARNING] Unsloth: The DAPO paper recommends setting `beta = 0.0` to remove the KL term - You have set it to {beta}.')\n"
" mask_truncated_completions = True\n"
" epsilon_high = 0.28\n"
" beta = 0.0\n"
"\n"
)
extra_args += check_dr_grpo
Expand Down