feat(t5gemma2): add Flash Attention 2 support #45868
feat(t5gemma2): add Flash Attention 2 support
97142297
fix(t5gemma2/fa2): correct decoder mask format and paged-attn detection
ac4bfe5c
fix(t5gemma2): regenerate modeling file from modular converter to fix…
101b9e98
vasqu
commented
on 2026-05-11
vasqu
commented
on 2026-05-11
Assignees
No one assigned
Login to write a write a comment.
Login via GitHub