microsoft/onnxruntime

Pull Requests Commits

Add WinML-focused documentation and DirectML deprecation notices

Copilot committed 159 days ago

3263cec4

Copilot committed 159 days ago

68a8cfcf

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

Craigacp committed 159 days ago

Verified c9bdbd70

[OVEP] OpenVINO EP Features and bug-fixes for ORT-1.23 (#25884)

preetha-intel committed 159 days ago

Verified 7a919c69

Add custom ops library_path to EP metadata (#25830)

psakhamoori committed 159 days ago

Verified ca77b7ed

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

qti-yuduo committed 160 days ago

Verified 3fc9779d

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

qti-hungjuiw committed 160 days ago

Verified 4754a1d6

Language bindings for model compatibility API (#25878)

adrastogi committed 160 days ago

Verified 820554e2

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

chilo-ms committed 160 days ago

Verified 1eb18f18

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

bachelor-dou committed 160 days ago

Verified 47f355aa

python GPU IO Bindings for NVIDIA (#25776)

ishwar-raut1 committed 160 days ago

Verified 3563f2e5

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

chwarr committed 160 days ago

Verified 179f371f

Fix MoE CPP tests (#25877)

apsonawane committed 160 days ago

Verified abe485ee

Fixes for DynamicQuantizeMatMul and Attention3D tests (#25814)

JonathanC-ARM committed 161 days ago

Verified 574806bf

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

derdeljan-msft committed 161 days ago

Verified e525ea22

DequantizeLinear should support non-zero zero_point when input type is int32 (#25646)

wangw-1991 committed 161 days ago

Verified 0b152002

Improve SimplifiedLayerNorm by using same techniques as SkipSimplifiedLayerNorm (#25850)

sushraja-msft committed 161 days ago

Verified d4e31dc6

[WebNN] Support Round op (#25810)

Honry committed 161 days ago

Verified 1d07e94c

[WebNN] Fix the op support limit for batchNormalization (#25856)

shiyi9801 committed 161 days ago

Verified 7af42b8f

Add API for precompiled model compatibility check using just the compat info (#25841)

adrastogi committed 161 days ago

Verified 3cadbdb4

Bump actions/setup-java from 4 to 5 (#25840)

dependabot[bot] committed 161 days ago

Verified 0ba4d296

Add default constructor to Ort::Status. (#25860)

yuslepukhin committed 161 days ago

Verified c9ec1da4

Fix a long standing bug on file memory mapping on windows. (#25833)

yuslepukhin committed 161 days ago

Verified 568ad202

[QNN-EP] Enable einsum with QK equations for QNN. (#25861)

quic-muchhsu committed 161 days ago

Verified c9c23b0e

Add cuda graph implementation for NV TRT RTX EP (#25787)

umangb-09 committed 162 days ago

Verified 16ae99ed

[webgpu] Optimize dp4 prefill shader for Qualcomm (#25578)

qjia7 committed 162 days ago

Verified 7e3174b0

[webgpu] Expand Unsqueeze version to 23 (#25858)

qjia7 committed 162 days ago

Verified f58f7eb7

Enable ABSL_FLAGS flag registration for onnxruntime_perf_test for mobile build (#25849)

chilo-ms committed 162 days ago

Verified 0de1c01e

[CPU] Improve QMoE kernel (#25822)

apsonawane committed 162 days ago

Verified db4b0f4e

Fix keyboard navigation accessibility for DocFX tab controls (#25819)

Copilot committed 162 days ago

Verified ce906093

Older