Skip to content

Conversation

@joecummings
Copy link
Member

@casteryh Let's consolidate

@meta-cla meta-cla bot added the CLA Signed This label is managed by the Meta Open Source bot. label Nov 5, 2025

# Dataset configuration
dataset:
path: "open-thoughts/OpenThoughts3-1.2M"
Copy link
Contributor

Choose a reason for hiding this comment

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

This seems like a dataset for SFT? Is the thinking part used for training? Or we only use question & answer?

Copy link
Member Author

Choose a reason for hiding this comment

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

Thinking part - sft is a pre-step for odp

**kwargs,
) -> torch.Tensor:
student_logprobs = compute_logprobs(logits, response)
reverse_kl = -(student_logprobs - teacher_logprobs)
Copy link
Contributor

Choose a reason for hiding this comment

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

Suggested change
reverse_kl = -(student_logprobs - teacher_logprobs)
reverse_kl = -(student_logprobs.detach() - teacher_logprobs)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

CLA Signed This label is managed by the Meta Open Source bot.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants