Skip to content
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.

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

data: fix the math dataset source #175

Open
wants to merge 3 commits into
base: main
Choose a base branch
from

Conversation

eric-haibin-lin
Copy link
Collaborator

since 'lighteval/MATH' is no longer available on huggingface.

@@ -397,6 +397,10 @@ def compute_ref_log_prob(self, data: DataProto):
torch.cuda.empty_cache()
return output

@register(dispatch_mode=Dispatch.MEGATRON_COMPUTE_PROTO)
def compute_log_prob(self, data: DataProto):
raise NotImplementedError()
Copy link
Collaborator

Choose a reason for hiding this comment

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

To fix the missing compute_log_prob in megatron-lm backend, I think we can simply return an empty dataProto and print some warning.

As the generate sequences func in megatron_workers already recompute the log_prob.
See: https://github.com/volcengine/verl/blob/main/verl/workers/megatron_workers.py#L368-L373

Copy link
Collaborator Author

Choose a reason for hiding this comment

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

I'll fix it soon

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants