Skip to content

[algo] fix: normalize GDPO advantages over responses #6497

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
lucky9-cyou wants to merge 1 commit into
base: main
Choose a base branch
from
Open

[algo] fix: normalize GDPO advantages over responses #6497

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
6 changes: 5 additions & 1 deletion verl/trainer/ppo/core_algos.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -463,7 +463,11 @@ def compute_gdpo_outcome_advantage(
else:
new_advantage += weights[i] * normalized_score

advantages = verl_F.masked_whiten(new_advantage, response_mask) * response_mask
response_level_advantage = verl_F.masked_mean(new_advantage, response_mask, axis=-1)
response_level_mask = response_mask.sum(dim=-1) > 0
response_level_advantage = verl_F.masked_whiten(response_level_advantage, response_level_mask)
Comment on lines +466 to +468
Copy link
Copy Markdown
Contributor

@gemini-code-assist gemini-code-assist Bot May 27, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

critical

If the batch size is 1, or if only one sequence in the batch has a valid response (i.e., response_level_mask.sum() is 1), verl_F.masked_whiten will call masked_var which raises a ValueError: The sum of the mask is one, which can cause a division by zero. This will crash the training or validation loop.

To prevent this, we should check if response_level_mask.sum() > 1 before applying masked_whiten. If there is at most one valid sequence, we can safely set the whitened advantages to zero. Additionally, converting response_level_mask to the same dtype as response_level_advantage ensures compatibility across different PyTorch versions and hardware backends.

Suggested change
response_level_advantage = verl_F.masked_mean(new_advantage, response_mask, axis=-1)
response_level_mask = response_mask.sum(dim=-1) > 0
response_level_advantage = verl_F.masked_whiten(response_level_advantage, response_level_mask)
response_level_advantage = verl_F.masked_mean(new_advantage, response_mask, axis=-1)
response_level_mask = (response_mask.sum(dim=-1) > 0).to(response_level_advantage.dtype)
if response_level_mask.sum() > 1:
response_level_advantage = verl_F.masked_whiten(response_level_advantage, response_level_mask)
else:
response_level_advantage = torch.zeros_like(response_level_advantage)


advantages = response_level_advantage.unsqueeze(-1) * response_mask

return advantages, advantages

Expand Down