onnxruntime
[webgpu] Apply Flash Attention if sliding window exceeds KV cache length
#25594

Merged

[webgpu] Apply Flash Attention if sliding window exceeds KV cache length #25594

guschmue merged 5 commits into microsoft:main from daijh:supports-sliding-window-for-flash-attention

Apply flash attention if sliding window exceeds KV cache length

7a827847

Fix typo

ade51b09

Check sequence length

5f00414c

qjia7 commented on 2025-07-30

guschmue added ep:WebGPU

qjia7 commented on 2025-07-31

Resolve comments

52ce1a14

qjia7 dismissed these changes on 2025-07-31

daijh dismissed their stale review via 4153d115 150 days ago

Minor update comment

4153d115

guschmue approved these changes on 2025-07-31

guschmue merged 7cc93cf4 into main 149 days ago

daijh deleted the supports-sliding-window-for-flash-attention branch 148 days ago

Reviewers

guschmue

qjia7

Assignees

No one assigned

Labels

ep:WebGPU

Milestone

No milestone