Skip to content

tch fix_bug in grpo_trainer.py #23

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

tch fix_bug in grpo_trainer.py #23

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 src/open_r1/trainer/grpo_trainer.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -415,10 +415,10 @@ def get_per_token_logps(model, input_ids, **kwargs):

with torch.inference_mode():
if self.ref_model is not None:
ref_per_token_logps = get_per_token_logps(self.ref_model, prompt_completion_ids)
ref_per_token_logps = get_per_token_logps(self.ref_model, prompt_completion_ids, **prompt_inputs) # tch fix_bug
else:
with self.accelerator.unwrap_model(model).disable_adapter():
ref_per_token_logps = get_per_token_logps(model, prompt_completion_ids)
ref_per_token_logps = get_per_token_logps(model, prompt_completion_ids, **prompt_inputs) # tch fix_bug
ref_per_token_logps = ref_per_token_logps[:, prompt_length - 1 :]

# Compute the KL divergence between the model and the reference model
Expand Down