Skip to content

train_continue resumes weights only (optimizer state not restored), may causing performance degradation #795

@wr0124

Description

@wr0124

When resuming training with the train_continue flag, only the model weights are uploaded/restored. The optimizer state is not preserved. As a result, training does not truly resume from the previous checkpoint and may lead to unstable updates and noticeable performance degradation after continuing training.

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