transformers
6daa3eeb
- Fix InternVL attention when using qk_norm (38B and 78B) (#37620)
Go
Login via GitHub
Home
Pricing
FAQ
Install
Login
via GitHub
Commit
View On
GitHub
Commit
352 days ago
Fix InternVL attention when using qk_norm (38B and 78B) (#37620) * fix internvlvision attention when using qk_norm * nit * modular
References
v4.51.3-InternVL-release
#37620 - Fix InternVL attention when using qk_norm (38B and 78B)
Author
yonigozlan
Parents
27a25bee
Loading