microsoft/onnxruntime

Pull Requests Commits

address comments

apsonawane committed 236 days ago

5c04dc95

Address comments

apsonawane committed 237 days ago

4bfc5089

QMoE kernel further optimizations

apsonawane committed 237 days ago

cb6939de

[CPU] Block-wise QMoE kernel for CPU (#26009)

apsonawane committed 237 days ago

Verified 5d177345

[WebNN] Support GatherBlockQuantized op (#26020)

Honry committed 238 days ago

Verified 203e2b26

Add opset imports for internal domains to Model Editor model (#26015)

skottmckay committed 238 days ago

Verified bd28856f

Update GitHub Actions to latest versions (#26022)

Changming Sun committed 240 days ago

Verified b6172c08

Integrate SME1 SGEMM KleidiAI kernels (#25760)

patryk-kaiser-ARM committed 240 days ago

Verified ec3bf7f0

[QNN] Fix multiple issues with per_channel quantization of QNN models (#25986)

qti-kromero committed 241 days ago

Verified f418a44c

[QNN EP] Fix GatherND int64 unique name creation (#26014)

quic-ashwshan committed 241 days ago

Verified 9aa3423d

Add ONNX patch to avoid regenerating proto files and improve rebuild time. (#26011)

edgchen1 committed 242 days ago

Verified 0b96cbef

[QNN EP] Upgrade QNN to 2.38.0 (#26013)

qti-jkilpatrick committed 242 days ago

Verified 953faa7c

[WebNN] Fix bug in quantizeLinear and dequantizeLinear (#25931)

Honry committed 242 days ago

Verified 99ae86b5

[CPU] MoE Kernel (#25958)

apsonawane committed 242 days ago

Verified 930e6408

[QNN] Relax tolerances for ONNX attention node tests (#26004)

qti-kromero committed 242 days ago

Verified 9ca0d69a

[Web] Avoid unnecessary data copy for pre-allocated tensors (#25571)

Honry committed 242 days ago

Verified 7fa13a68

Check for non-empty ep_nodes before resizing index_to_ep_node in EpGraph::CreateImpl(). (#26001)

edgchen1 committed 243 days ago

Verified 211d5fa2

Update React Native iOS CI build. (#25967)

edgchen1 committed 243 days ago

Verified e3ea71a9

Pin cmake version in macOS github Actions (#25998)

Changming Sun committed 243 days ago

Verified e11a20c6

Build break fix for NV TRT RTX EP (#25991)

umangb-09 committed 243 days ago

Verified ba7ce768

Bump ruff from 0.12.9 to 0.12.12 (#25975)

dependabot[bot] committed 243 days ago

Verified 50241677

Bump actions/setup-python from 5 to 6 (#25979)

dependabot[bot] committed 243 days ago

Verified e689098f

Extend ORT perf test to report submit vs inference time (#25895)

gedoensmax committed 243 days ago

Verified 7434de31

Address a bug when a number of meminfos return by SessionGetOutputCount() always for inputs (#25988)

yuslepukhin committed 243 days ago

Verified 753e6c5c

Remove CACHE_URL settings from Github Actions (#25989)

Changming Sun committed 243 days ago

Verified 92a5b040

Fix Mac Catalyst build options. (#25970)

edgchen1 committed 243 days ago

Verified 98767b05

[WebGPU] Remove an unreachable return in matmul_nbits.cc (#25981)

Jiawei-Shao committed 244 days ago

Verified 0ca319b3

Revert accidental memory alignment change for x86 for prepacked weights buffer (#25971)

hariharans29 committed 244 days ago

Verified 96f45950

Integration with ONNX 1.19 (#25678)

titaiwangms committed 247 days ago

Verified ecb26fb7

[QNN EP] fix memory leak in OrtEpFactory::GetSupportedDevices() (#25968)

adrianlizarraga committed 247 days ago

Verified 8570298d

Older