Commits microsoft/onnxruntime

Merge remote-tracking branch 'origin/main' into copilot/fix-sigmoid-transformation-bug

Copilot committed 3 days ago

Verified be61557c

Fix CUDA Attention dispatch: skip MEA when head_size != v_head_size in GQA (#28358)

justinchuby committed 3 days ago

Verified 1f257837

chore: rename `ort_api_1_to_26` to `ort_api_1_to_27` (#28341)

sanaa-hamel-microsoft committed 3 days ago

Verified ef446045

fix(ci): test pipeline didn't correctly specify `ReleaseVersionSuffix` (#28346)

sanaa-hamel-microsoft committed 3 days ago

Verified ebee6069

fix: make sympy an optional runtime dependency (#28141)

Rishi-Dave committed 4 days ago

Verified b81f3f85

Fix ReshapeFusion dropping allowzero on inferred 0-sized intermediate dims (#28349)

titaiwangms committed 4 days ago

Verified 7529033f

Suppress -Wmaybe-uninitialized for onnxruntime_pybind11_state under pybind11 3.0 (#28251)

Rishi-Dave committed 4 days ago

Verified 8a095016

adds foundry local packaging to webgpu plugin ep packaging pipeline (#28300)

prathikr committed 4 days ago

Verified a1aa3bbf

Fix BitShift UB when shift amount >= bit width (#28272)

Copilot committed 4 days ago

Verified 3454f86e

Eliminate Legacy MHA Unfused path from ONNX Attention; unify on 3-tier dispatch with causal alignment fix (#27992)

titaiwangms committed 4 days ago

Verified 4ca6b228

Add plugin-ep-webgpu/RELEASE.md (#28321)

edgchen1 committed 4 days ago

Verified 40c9f85f

refactor(ci): simplify build date/time metadata propagation (#28294)

sanaa-hamel-microsoft committed 4 days ago

Verified f1c96d56

Bump plugin-ep-webgpu/VERSION_NUMBER to 0.2.0. (#28322)

edgchen1 committed 4 days ago

Verified 96786efa

Bump version for 1.27.0 (#28324)

eserscor committed 4 days ago

Verified b81e0c65

Support ONNX overloaded functions (IR version 10+) (#28275)

gramalingam committed 5 days ago

Verified 763c4e35

docs: add opset version notation legend to OperatorKernels.md (#28143)

Rishi-Dave committed 6 days ago

Verified 95828673

Add option to memory map .ORT model loads (#28164)

Kevin-Taha committed 7 days ago

Verified 9d1492a4

add support for DFT with onesided=True and inverse=True (irfft) (#27028)

simonbyrne committed 7 days ago

Verified 1e4ee666

GridSample: harden float->int64 casts against NaN/Inf/out-of-range coords (#28302)

GopalakrishnanN committed 7 days ago

Verified 55c5c822

Relax GQA seqlens_k shape validation for backward compat with older models (#28259)

vraspar committed 7 days ago

Verified 60ce9ccd

Fix DoubleQDQPairsRemover adding spurious dimension to scalar scale/zero-point (#28057)

ssam18 committed 7 days ago

Verified d02a0fd5

remove weights_are_all_positive_ from TreeEnsemble (#27552)

xadupre committed 7 days ago

Verified 9b30f306

fix(ci): incorrect relative template includes for setup-feeds (#28312)

sanaa-hamel-microsoft committed 7 days ago

Verified 5f2f848a

Add QNN Plugin EP repo link to README (#28225)

vraspar committed 7 days ago

Verified de2bc90c

Include license file in built distributions (#27783)

julia-thorn committed 7 days ago

Verified 8dd4a068

Fix CUDA 13 build error in gqa_unfused_attention.cu (#28309)

tianleiwu committed 7 days ago

Verified 6e19374b

[OVEP] OpenVINO EP 1.26.0 Development Release Updates (#28297)

ankitm3k committed 7 days ago

Verified d6c363c0

Fix NVTX profiling: restore missing node_compute_range_.Begin() call (#28273)

Copilot committed 8 days ago

Verified 76655a03

Fix session logger use-after-free during EP teardown under VERBOSE logging (#28274)

Copilot committed 8 days ago

Verified 5f7e9d0c

Implement CUDA EP Plugin profiling API (#28216)

yuslepukhin committed 8 days ago

Verified 9ca299d0