Skip to content

fix bug of GRPO default setting #113

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

Open
MozerWang wants to merge 1 commit into
base: main
Choose a base branch
from
Open

fix bug of GRPO default setting #113

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

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 2 additions & 2 deletions examples/train_oat/train_oat_grpo.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -111,7 +111,7 @@ class Args(PPOArgs):

# Reward settings
gamma: float = 1.0 # Discount factor for Monte Carlo returns
norm_return: bool = True
norm_return: bool = False # Should be False to avoid double normalization

# online evaluation settings
eval_envs: str = None # 'eval:AIME24|eval:MATH500'. See gem.envs
Expand All @@ -126,7 +126,7 @@ class Args(PPOArgs):
# Episode collection logic
keep_generation_failed: bool = False # Keep episodes with generation failures

critic_type2: Literal["grpo", "drgrpo", "rloo", "ep_level", "none"] = "none"
critic_type2: Literal["grpo", "drgrpo", "rloo", "ep_level", "none"] = "grpo"


""" +=======================================+ """
Expand Down