onnxruntime
Validate Attention projection and cache dimensions
#32946

Merged

Validate Attention projection and cache dimensions #32946

GopalakrishnanN merged 7 commits into microsoft:main from GopalakrishnanN:gnallasamy-microsoft-glowing-guacamole

Validate Attention projection and cache dimensions

3e2cccd4

GopalakrishnanN requested a review from

copilot-pull-request-reviewer 31 days ago

copilot-pull-request-reviewer commented on 2026-09-29

Refresh Android minimal binary-size baseline

6792da7c

Skip unavailable WebGPU Attention test providers

b313a38b

tianleiwu commented on 2026-10-05

Refresh Attention validation and preserve main prepacking diagnostics…

f49b490e

Refresh PR #32946 with main CUDA integration build fix

7a636379

Assert preserved nonpositive QKV diagnostics in shared validation cases

1b749d2e

GopalakrishnanN requested a review from

tianleiwu 23 days ago

Refresh #32946 with main node-loader CI baseline fix

b3097c09

tianleiwu approved these changes on 2026-10-21

GopalakrishnanN closed this 9 days ago

GopalakrishnanN reopened this 9 days ago

GopalakrishnanN merged 2e912580 into main 8 days ago

Reviewers

tianleiwu

Assignees

No one assigned

Labels

None yet

Milestone

No milestone