Skip to content

v1.1.0

Choose a tag to compare

@martinloretzzz martinloretzzz released this 07 Oct 09:11
· 42 commits to main since this release
2c2a100
  • Adds support for the NX-AI/TiRex-1.1-gifteval model
  • Improved torch.compile speed. Enable by setting compile=True on load_model
  • Add frequency depended upsampling