Skip to content

In evaluation, different models come up with different sizes of test dataset #63

Description

@alpha-tango

This is in the performing inference logging step:

  • HumpbackNorthAtlantic: 884
  • perch_v2: 757

Actual number in combined_annotations_full:

grep 'eval' ../combined_annotations_full/MyNewTFRecordsDataset/combined_annotations.csv | wc -l
12126

I haven't changed the batch_size away from 32, so it's surprising to me that the numbers aren't the same.

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