Skip to content

problem about train process #6

@lingcon01

Description

@lingcon01
Image Thank you for your excellent work. However, I encountered the following issue during training: I am using your code as well as the dataset released with TargetDiff. The loss for positions remains very high and does not decrease. Could you please let me know what might be causing this problem?

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