Support direct quantization for FP8 matmul #3922
copybara-service
merged
4 commits into
google:main
from
wenscarl:direct_quant
wenscarl
changed the title 
kaixih
commented
on 2024-05-23
wenscarl
changed the title
kaixih
approved these changes
on 2024-08-30
Direct quantization for FP8 Dense Layer.
6f5cee12
kaixih
approved these changes
on 2024-09-03
levskaya
approved these changes
on 2024-09-04
levskaya
approved these changes
on 2024-09-04
Clean imports
4f5a722e
Add type: ignore
b4dc0949
Remove unnecessary warnings.
e9d19aca
Login to write a write a comment.
Login via GitHub