mirror of
https://github.com/lucidrains/DALLE2-pytorch.git
synced 2026-02-23 07:44:40 +01:00
switch to using linear attention for the sparse attention layers within unet, given success in GAN projects
This commit is contained in: