microsoft/onnxruntime

Pull Requests Commits

Disable quantized Mlas, still not giving good tps

apsonawane committed 107 days ago

d57a7c3e

Add overflow safety changes

apsonawane committed 108 days ago

8289fcb1

Re-enable quantized Mlas

apsonawane committed 108 days ago

d399c66d

Fix merge conflicts

apsonawane committed 108 days ago

83ffc5d9

[CPU] Fix compilation errors because of unused variables (#26147)

apsonawane committed 108 days ago

Verified 42fcd71a

[QNN-EP] Fix logic flow bug (#26148)

quic-calvnguy committed 108 days ago

Verified b160e8c6

[QNN-EP] Enable largest attr and remove sorted check for topk. (#26150)

quic-muchhsu committed 108 days ago

Verified b608f798

MultiheadAttention CUDA BF16 Support (#26083)

nenad1002 committed 108 days ago

Verified b7ae53f7

[webgpu] Manually normalize `Conv-Transpose` dispatch group size for Intel GPUs (#25908)

daijh committed 108 days ago

Verified f2f50ebc

Add BF16 kernels in several ops for Gemma-3 (#26102)

kunal-vaishnavi committed 108 days ago

Verified 99e10e68

[QNN EP] Add support for RandomUniformLike Op in QNN EP (#26123)

quic-ashwshan committed 108 days ago

Verified 8c5d2b58

Declare Shape, Reshape, Transpose, Squeeze, Unsqueeze for opsets 21, 23 on CUDA (#26075)

xadupre committed 108 days ago

Verified 6650e07b

Regenerate test model with ONNX IR < 12 (#26149)

adrianlizarraga committed 109 days ago

Verified 88f2652d

[C#] Implement missing APIs (#26101)

yuslepukhin committed 109 days ago

Verified 35dcab50

[CUDA] Fix cuda 12.8 build warnings (#26136)

tianleiwu committed 109 days ago

Verified ad32f424

[TRT RTX EP] Fix bug for generating the correct subgraph in GetCapability (#26132)

chilo-ms committed 110 days ago

Verified 72e56e76

[CORE]Fix windows warning with CUDA fp4 header inclusion (#26134)

hariharans29 committed 110 days ago

Verified bdffd76c

MemcpyFromHost and MemcpyToHost support for plugin EPs (#26088)

edgchen1 committed 110 days ago

Verified 45457323

[WebGPU] Fix potential dangling reference in MatMul (#26126)

xiaofeihan1 committed 110 days ago

Verified ff66c70b

[Natvis] Improve InlinedVector rendering (#26124)

yuslepukhin committed 111 days ago

Verified 1c436de5

Delete python 3.10 from packaging pipeline (#26100)

Changming Sun committed 111 days ago

Verified 9140a985

[ARM CPU] SVE support for Elementwise kernels (#25238)

sanketkaleoss committed 111 days ago

Verified 79a2d250

Fix CUDA 12.8+ build issues with regards to fp4 type (#26061)

hariharans29 committed 111 days ago

Verified 99ee627d

Bump actions/setup-python from 5 to 6 (#26113)

dependabot[bot] committed 111 days ago

Verified a7e55b02

Fix compilation errors with TRT-RTX EP (#26097)

gaugarg-nv committed 111 days ago

Verified 0a65dd1b

Convert React Native pipeline to GH (#26052)

Changming Sun committed 114 days ago

Verified 2cc77384

[QNN-EP] Add GPU to QNN EP factory (#26085)

qti-mattsinc committed 114 days ago

Verified a7ae9ae7

[QNN EP] Add Support for STFT Op in QNN EP (#26063)

quic-ashwshan committed 114 days ago

Verified d3a916db

Add model compilation in ORT perf test (#25797)

thevishalagarwal committed 114 days ago

Verified cced33b2

[js] batch npm audit fix (#26094)

fs-eire committed 115 days ago

Verified d3096cd8

Older