[qwen2 vl] fix packing with all attentions #39447
fix qwen2 vl packing in FA2
048c3756
why? delete!
b200d1f4
qwen2-5-vl seems to work now
48e4a176
update
5ba588b0
Merge branch 'main' into qwen2-vl-fa2
84091a30
fix tests
942e7b25
Merge branch 'main' into qwen2-vl-fa2
81068aed
start by adapting FA2 tests
9f486aa8
add similar tests for sdpa/eager
786f5f6b
address comments
58026ded
zucchini-nlp
changed the title [qwen2 vl] fix packing with FA2 [qwen2 vl] fix packing with all attentions 245 days ago
Merge branch 'main' into qwen2-vl-fa2
5aea0137
why is this even in conditional model and not base model?
e218c6e7
Merge branch 'main' into qwen2-vl-fa2
23350ec2
Merge branch 'main' into qwen2-vl-fa2
66a5116a
Assignees
No one assigned
Login to write a write a comment.
Login via GitHub