onnxruntime
[webgpu] Optimize Attention by enhancing flash attention support
#26715

Merged

[webgpu] Optimize Attention by enhancing flash attention support #26715

qjia7 merged 9 commits into main from attention_opt

[webgpu] Enable attention_bias

60203a10

support batch

80daa724

fix shader errors

39ea5c90

fix the incorrect results

f954abd1

fix CI errors

b1c192a3

fix errors in copykvcache

53a944a6

xenova approved these changes on 2025-12-04

fix the bots errors

73d0159c

remove unnecessary checking

5af305ed

missing

c2de85da

qjia7 marked this pull request as ready for review 255 days ago

qjia7 requested a review from

guschmue 255 days ago

qjia7 requested a review from

fs-eire 255 days ago

guschmue added ep:WebGPU

guschmue approved these changes on 2025-12-09

qjia7 merged 549d7415 into main 254 days ago

qjia7 deleted the attention_opt branch 254 days ago

Reviewers

guschmue

xenova

fs-eire

Assignees

No one assigned

Labels

ep:WebGPU

Milestone

No milestone