Skip to content

How to gurantee the output.logits.shape[:-1] == labels.shape #87

Description

@foreverhell

dpo
How to gurantee the two the same?
When I train a custom LLM in DPO, the loss cannot divergence. Is the reason for the two are different?

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions