Extend Attention Bias Broadcast Support #21710
broadcast attention_bias dim 0 and 1
91284f07
broadcast attn bias in decoder masked mha
c76f2940
tianleiwu
marked this pull request as draft 1 year ago
Add MHA tests
a8cebba1
rename relative_position_bias to attention_bias
c728b0be
fix build
2bff1881
update doc
58792dd5
Merge branch 'main' into tlwu/mha_attn_bias
801a86e3
format js
acfd6117
refactoring
1eb8c6b3
tianleiwu
marked this pull request as ready for review 1 year ago
refactoring cpu; add comments
6766b17e
refine softmax kernel
4984b455
benchmark mha with attention bias
a7e221b9
mark maybe_unused
1226c6d0
refine attn_bias_offset for dmmha with asummption of S=1
0c7f3952
jchen351
approved these changes
on 2024-08-16
tianleiwu
merged
d79e3c57
into main 1 year ago
tianleiwu
deleted the tlwu/mha_attn_bias branch 1 year ago
Assignees
No one assigned
Login to write a write a comment.
Login via GitHub