Sync with Microsoft ONNX Runtime - 02/09/2025#804

Closed

Jaswanth51 wants to merge 16 commits intoovep-developfrom

sync_msft_02092025

Commits on Sep 3, 2025

[CPU] Optimize GQA attention bias application for FP16 (#25871)

derdeljan-msft
authored and
Jaswanth51
committed
Fixes for DynamicQuantizeMatMul and Attention3D tests (#25814)

JonathanC-ARM
authored and
Jaswanth51
committed
Fix MoE CPP tests (#25877)

apsonawane
authored and
Jaswanth51
committed
[c++] Eliminate dynamic initialization of static Ort::Global<void>::api_ (#25741)

authored and
Jaswanth51
committed
python GPU IO Bindings for NVIDIA (#25776)

authored and
Jaswanth51
committed
[CANN] Add a `enable_cann_subgraph` feature parameter (#25867)

bachelor-dou
authored and
Jaswanth51
committed
[EP ABI] Add OpAttr_GetTensorAttributeAsOrtValue and replace the existing Node_GetTensorAttributeAsOrtValue (#25886)

chilo-ms
authored and
Jaswanth51
committed
Language bindings for model compatibility API (#25878)

authored and
Jaswanth51
committed
[QNN-EP] Introduce Level1 Transformer into qnn.preprocess (#25883)

qti-hungjuiw
authored and
Jaswanth51
committed
[QNN EP] Minor fix weight name missing when not valid QDQ node group (#25887)

qti-yuduo
authored and
Jaswanth51
committed
Add custom ops library_path to EP metadata (#25830)

psakhamoori
authored and
Jaswanth51
committed
[java] Auto EP and compile model support (#25131)

Craigacp
authored and
Jaswanth51
committed
Add error handling to extract_nuget_files.ps1 (#25866)

snnn
authored and
Jaswanth51
committed
[Fix] illegal memory access in GetInputIndices with optional inputs (#25881)

mingyueliuh
authored and
Jaswanth51
committed
Re-enable cpuinfo for ARM64EC (#25863)

edgchen1
authored and
Jaswanth51
committed
[webgpu] Add back missing code comments for flash decoding (#25879)

xiaofeihan1
authored and
Jaswanth51
committed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Sync with Microsoft ONNX Runtime - 02/09/2025#804

Sync with Microsoft ONNX Runtime - 02/09/2025#804
Jaswanth51 wants to merge 16 commits intoovep-developfrom
sync_msft_02092025

Commits on Sep 3, 2025

[CPU] Optimize GQA attention bias application for FP16 (#25871)

Fixes for DynamicQuantizeMatMul and Attention3D tests (#25814)

Fix MoE CPP tests (#25877)

[c++] Eliminate dynamic initialization of static Ort::Global<void>::api_ (#25741)

python GPU IO Bindings for NVIDIA (#25776)

[CANN] Add a `enable_cann_subgraph` feature parameter (#25867)

[EP ABI] Add OpAttr_GetTensorAttributeAsOrtValue and replace the existing Node_GetTensorAttributeAsOrtValue (#25886)

Language bindings for model compatibility API (#25878)

[QNN-EP] Introduce Level1 Transformer into qnn.preprocess (#25883)

[QNN EP] Minor fix weight name missing when not valid QDQ node group (#25887)

Add custom ops library_path to EP metadata (#25830)

[java] Auto EP and compile model support (#25131)

Add error handling to extract_nuget_files.ps1 (#25866)

[Fix] illegal memory access in GetInputIndices with optional inputs (#25881)

Re-enable cpuinfo for ARM64EC (#25863)

[webgpu] Add back missing code comments for flash decoding (#25879)

Sync with Microsoft ONNX Runtime - 02/09/2025#804Jaswanth51 wants to merge 16 commits intoovep-developintel/onnxruntime:ovep-developfrom sync_msft_02092025intel/onnxruntime:sync_msft_02092025Copy head branch name to clipboard

Commits

Commits on Sep 3, 2025

Sync with Microsoft ONNX Runtime - 02/09/2025#804
Jaswanth51 wants to merge 16 commits intoovep-developfrom
sync_msft_02092025