Skip to content

Releases: locuslab/convmixer

(ignore) timm weights

10 Oct 00:34
e026544
Compare
Choose a tag to compare
(ignore) timm weights Pre-release
Pre-release

These weights have slightly different parameter names and aren't compatible with this codebase.

ConvMixer Weights

09 Oct 18:13
Compare
Choose a tag to compare

We provide weights for:

  • ConvMixer-1536/20 (k = 9, p = 7)
  • ConvMixer-768/32 (k = 7, p = 7)
    • IMPORTANT: This model used ReLU instead of GELU.
    • Currently, you would need to change nn.GELU() to nn.ReLU() in convmixer.py to use these weights; we will fix this later.
  • ConvMixer-1024/20 (k = 9, p = 14)