Fix packed GatedDeltaNet shape inference #32708
kunal-vaishnavi
merged
49 commits into
main
from
fix-32645-gated-delta-net-packed-shape-inference
Add WebGPU GatedDeltaNet kernel
e626c13c
Fix optional WebGPU GatedDeltaNet state
228f537d
Validate WebGPU GatedDeltaNet inputs
66d3344a
Address WebGPU GatedDeltaNet review comments
064d1f74
Clarify WebGPU GatedDeltaNet state bindings
6d902419
Fix WebGPU GatedDeltaNet context ambiguity
b6554023
Fix GatedDeltaNet WGSL template parameters
79e2319c
Address GatedDeltaNet WebGPU review feedback
8b12dc1a
Fix segmented GatedDeltaNet WebGPU buffers
3c0031bf
Optimize GatedDeltaNet WebGPU dispatch
fcf7fc04
Add WebGPU packed buffer views and tiled GatedDeltaNet
8fe9e5b9
Pack large GatedDeltaNet WebGPU inputs
079fd492
Add GatedDeltaNet parallel prefill
b6046f3f
Fix GatedDeltaNet WGSL and segmentation test
76eb0061
Guard WebGPU-only segmentation test
d3a19455
Fix non-WebGPU GatedDeltaNet test build
7a2c8b56
Fix WebGPU packed-view cache key collisions
1150793e
Fix WebGPU shader offset literals
0d6207d6
Fix GatedDeltaNet alias test inputs
381857dd
Include WebGPU adapter definitions
6f1b4229
Fix WebGPU plugin API initialization mismatch
302e803f
Fix WebGPU plugin logger definitions
4039fdaa
Fix GatedDeltaNet parameter dispatch
f111a98e
Optimize GatedDeltaNet recurrent WebGPU execution
c92cedef
Cover optimized GatedDeltaNet recurrent path
c4b85d07
Exercise vectorized GatedDeltaNet path on baseline adapters
8fc4970a
Add effective WebGPU limit overrides
750a2aa1
Fix GatedDeltaNet WebGPU CI compilation
665cf2ec
Fix packed GatedDeltaNet WebGPU test setup
3967488d
Address GatedDeltaNet WebGPU review feedback
ebc53528
Fix segmented WebGPU test provider setup
4de9f0f2
Exercise segmented aliased WebGPU state
4f36b362
Fix packed GatedDeltaNet shape inference
c734303c
Support packed QKV in CUDA GatedDeltaNet
e58a7dc5
Support packed QKV in WebGPU GatedDeltaNet
6186e341
Merge remote-tracking branch 'origin/main' into copilot/32510-perform…
34db1187
Merge remote-tracking branch 'origin/main' into copilot/32510-perform…
7f44c334
Address GatedDeltaNet packed-QKV review feedback (#33013)
9fb8f3b0
Merge remote-tracking branch 'origin/main' into copilot/32510-perform…
922d8d42
Merge target branch and resolve GatedDeltaNet shape inference conflict
5e8d1262
Update generated contrib operator documentation
0f7f1fad
Merge target updates and synchronize generated GatedDeltaNet docs
d555fa14
Merge remote-tracking branch 'origin/main' into copilot/32510-perform…
d4ad885f
Fix Pool1D test indexing for WASM
706f49c7
Merge latest target branch changes into packed GatedDeltaNet PR
9afed758
Base automatically changed from
copilot/32510-performance-optimizations
to
main
15 days ago
Merge main and resolve GatedDeltaNet test helper conflict
43de8524
apsonawane
commented
on 2026-10-01
Reject non-divisible packed GatedDeltaNet uniform batches
9b7ce819
Fix packed GatedDeltaNet regression test formatting
3363162f
apsonawane
commented
on 2026-10-01
apsonawane
commented
on 2026-10-01
apsonawane
commented
on 2026-10-01
Validate packed GatedDeltaNet shapes against execution requirements
add763c2
sayanshaw24
approved these changes
on 2026-10-01
Assignees
No one assigned
Labels
None yet
Milestone
No milestone
Login to write a write a comment.
Login via GitHub