Skip to content

Replicating 800 epoch pretraining ColorMAE green segmentation results #3

Description

@tristanengst

Thanks again for the great work with ColorMAE! I'm trying to replicate the 800-epoch green-noise ColorMAE downstream result on ADE20K semantic segmentation, and not getting to the reference accuracy—only 47.91 instead of 49.18. Concretely, I'm following the SLURM script for semantic segmentation finetuning of the 300 epoch model, but with model weights converted from the 800-epoch pretraining with green noise. Any insight on if this script wouldn't generalize to the 800-epoch pretraining result, or if something else might be amiss?

Activity

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

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions