Skip to content

Further long-running trainings #68

@texhnolyze

Description

@texhnolyze

Summary

Currently, the factor which seems to have made the biggest difference in model performance seems to have been length of training.
For this reason, we want to run further long-running trainings also using the B-Human data.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    Projects

    Status

    In progress

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions