Skip to content

what is "avg_reward" in the RL data? #7

Open
Open
what is "avg_reward" in the RL data?#7

Description

@zhangchbin
zhangchbin
opened on Dec 19, 2025
No description provided.

Activity

  1. KemingWu commented on Dec 24, 2025

    @KemingWu
    KemingWu
    on Dec 24, 2025
    Collaborator

    Hi, "avg_reward" in the RL data refers to the pass rate calculated using Qwen3-VL-8B after 8 rollouts.

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

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions