PR #21710 Extend Attention Bias Broadcast Support

Extend Attention Bias Broadcast Support #21710

tianleiwu merged 14 commits into main from tlwu/mha_attn_bias

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

github-advanced-security commented on 2024-08-15

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

tianleiwu requested a review 1 year ago

tianleiwu requested a review from

fs-eire 1 year ago

tianleiwu requested a review from

wangyems 1 year ago

tianleiwu requested a review from

kunal-vaishnavi 1 year ago

tianleiwu requested a review from

yufenglee 1 year ago

wangyems commented on 2024-08-15

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

tianleiwu requested a review from

wangyems 1 year ago

kunal-vaishnavi approved these changes on 2024-08-16

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

Reviewers

kunal-vaishnavi

jchen351

wangyems

github-advanced-security

fs-eire

yufenglee

Assignees

No one assigned

Labels

None yet

Milestone

No milestone

onnxruntime Extend Attention Bias Broadcast Support #21710 Merged

Extend Attention Bias Broadcast Support #21710

onnxruntime
Extend Attention Bias Broadcast Support
#21710

Merged