feat: enable head_dim=256
for attention kernels (#132)
#35
release_please.yml
on: push
release-please
18s
Matrix: wheel / build
Waiting for pending jobs
wheel
/
release