PR #25156 Attention Operator (CPU)

Attention Operator (CPU) #25156

xadupre merged 56 commits into main from xadupre/attention

Skeleton for Attention Operator (CPU)

39ff7993

github-actions commented on 2025-06-24

github-advanced-security commented on 2025-06-24

Update onnxruntime/core/providers/cpu/llm/attention.cc

ec9348ad

Merge branch 'main' of https://github.com/microsoft/onnxruntime into …

7ea6cbc1

first draft for attention

2174ab61

First working attention scenario

d03321ca

fix build issues

a6d803b3

fix with mask

72d8213b

add mask

ad946b7b

new addition

deaa572a

titaiwangms requested a review from

titaiwangms 195 days ago

add test on causal

00fa987c

improve kernel

929bd731

fix 3D

a8d19b99

Merge branch 'main' of https://github.com/microsoft/onnxruntime into …

da457b95

add softcap

d6c6c46c

fix present_value

8c2a53e0

fix gqa

bdbb773b

Fix gda past, present

88a04a34

fix qkmode

3d07fecf

add more unit test

099ba9a7

add ort_enforce

d45fd010

more enforce

08b76dac

xadupre changed the title ~~[DRAFT] Attention Operator (CPU)~~ Attention Operator (CPU) 187 days ago

xadupre marked this pull request as ready for review 187 days ago

titaiwangms requested a review from

tianleiwu 187 days ago

justinchuby requested a review from

justinchuby 187 days ago

justinchuby requested a review from

copilot-pull-request-reviewer 187 days ago

justinchuby added release:1.23.0

copilot-pull-request-reviewer commented on 2025-07-08

gramalingam commented on 2025-07-08

tianleiwu commented on 2025-07-08

gramalingam commented on 2025-07-08

tianleiwu commented on 2025-07-08

gramalingam commented on 2025-07-08

tianleiwu commented on 2025-07-08

titaiwangms commented on 2025-07-08

titaiwangms commented on 2025-07-09

Update onnxruntime/core/providers/cpu/llm/attention.cc

f3ccc4f8

Update onnxruntime/core/providers/cpu/llm/attention.cc

7567629d

address PR comments

bb53f0fb

justinchuby requested a review from

copilot-pull-request-reviewer 186 days ago

copilot-pull-request-reviewer commented on 2025-07-09

improve for 2D mask

6457762c

comment

e76c56b1

refactor

ac719af3

titaiwangms requested a review from

kunal-vaishnavi 185 days ago

useless variabel

6f45bb57

Merge branch 'main' of https://github.com/microsoft/onnxruntime into …

66109dbb

fix parameters

a58f41e6

kunal-vaishnavi commented on 2025-07-11

Update onnxruntime/core/providers/cpu/llm/attention_helper.h

3d9d5431

Merge branch 'main' of https://github.com/microsoft/onnxruntime into …

47c32044

github-actions commented on 2025-07-15

github-advanced-security commented on 2025-07-15

Update onnxruntime/core/providers/cpu/llm/attention_helper.h

e1e9ec57

Merge branch 'xadupre/attention' of https://github.com/microsoft/onnx…

c1b4e120

add support for float16

2dfb5d4d

enable gemm for float16 in Attention

34262fb6

support for float16

e31b42d0

cast

eb6eef07

titaiwangms requested a review from

tianleiwu 180 days ago

titaiwangms requested a review from

kunal-vaishnavi 180 days ago

tianleiwu commented on 2025-07-15

pr comments

bcc7c8a4

tianleiwu commented on 2025-07-16

fix input transposition

5fd5ac89

fix gemm fp16

5c89fe2b

fix issues

14786ad7

fix an error message

1a2f5fc2

enables removed tes

876c8eff

fix fp16 implementation

6b393ba1

Merge branch 'main' of https://github.com/microsoft/onnxruntime into …

94466597

fix missing cast

5b826d30

disable one warning

1035cf8c

disable attention 3d tests

1456623a

update test cases

351c3036

remove unnecessary comment

2b936168

Merge branch 'main' of https://github.com/microsoft/onnxruntime into …

3bfc049b

disable two tests

c7eb814c

Merge branch 'main' of https://github.com/microsoft/onnxruntime into …

9cf5712e

justinchuby dismissed these changes on 2025-07-22

kunal-vaishnavi commented on 2025-07-22

Update onnxruntime/core/providers/cpu/llm/attention.cc

45f5556b

xadupre dismissed their stale review via 45f5556b 173 days ago

xadupre requested a review from

tianleiwu 173 days ago

Merge branch 'main' of https://github.com/microsoft/onnxruntime into …

753dbbaa

titaiwangms assigned

titaiwangms 171 days ago

justinchuby approved these changes on 2025-07-24

titaiwangms approved these changes on 2025-07-24

tianleiwu approved these changes on 2025-07-25

xadupre merged c3499d78 into main 170 days ago

xadupre deleted the xadupre/attention branch 170 days ago

snnn removed release:1.23.0

Reviewers

tianleiwu

titaiwangms

justinchuby

kunal-vaishnavi

gramalingam

github-actions

github-advanced-security

copilot-pull-request-reviewer

Assignees

titaiwangms

Labels

None yet

Milestone

No milestone

onnxruntime Attention Operator (CPU) #25156 Merged

Attention Operator (CPU) #25156

onnxruntime
Attention Operator (CPU)
#25156

Merged