Skip to content

Conversation

@rwightman
Copy link
Collaborator

  • Add RandAugment to existing AutoAugment, ImageNet training supported, but not the RA/AA searches
  • Update a few training args, drop_connect, float epoch decay, etc
  • Bring in the JIT version of swish activation for EfficientNet separately from my WIP feature extraction because gains are significant

@rwightman rwightman merged commit db04677 into master Nov 22, 2019
@rwightman rwightman deleted the randaugment branch February 21, 2020 19:53
guoriyue pushed a commit to guoriyue/pytorch-image-models that referenced this pull request May 24, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants