microsoft/onnxruntime

Pull Requests Commits

Add CUDA architecture 120 support for RTX 5090 GPUs

Copilot committed 80 days ago

54a29724

Copilot committed 80 days ago

3f510380

Fix Local Attention off by 1 bug (#25927)

aciddelgado committed 81 days ago

Verified 8af9f580

Fix CPU EP Tile 0D overvalidation (#25821)

fdwr committed 81 days ago

Verified 978bfca5

[WebNN] Rename outputs for gru and lstm (#26196)

Honry committed 82 days ago

Verified 88b63d04

[MLAS] Fix SVE enabled pipeline break (#26203)

hariharans29 committed 83 days ago

Verified 189e673d

Normalize to positive axis for comparison on GEMM (#26152)

qti-yuduo committed 83 days ago

Verified 3ec8829d

Move KleidiAI support detection to CMake logic. (#26178)

edgchen1 committed 83 days ago

Verified 2e8a45ab

Fix(build): Correct maybe-uninitialized and range-loop-construct warnings (#26201)

Changming Sun committed 83 days ago

Verified 8ebd0bf1

[WebNN] Fix bug in int64 data type support validation (#26169)

Honry committed 83 days ago

Verified 710856f8

Add build option for ARM NCHWc kernels (#26171)

hariharans29 committed 83 days ago

Verified 04386c92

[CUDA] cutlass upgrade to 4.2.1 (#26154)

tianleiwu committed 83 days ago

Verified 048aaee6

Fix doxygen error (#26183)

yuslepukhin committed 86 days ago

Verified e5678a13

Support Optional shapes (#26164)

yuslepukhin committed 86 days ago

Verified 149d2089

Fix Windows non-CUDA build break (#26168)

hariharans29 committed 86 days ago

Verified 6b81b5f6

[QNN EP] Add dynamic option to set HTP performance mode (#26135)

quic-tirupath committed 86 days ago

Verified 6cc40fdb

Bump actions/download-artifact from 4 to 5 (#26112)

dependabot[bot] committed 87 days ago

Verified 0de21a76

[EP ABI] Check if nodes specified in GetCapability() have already been assigned (#26156)

adrianlizarraga committed 87 days ago

Verified 67d3ba0e

MLAS: use vmlaq_f32 for ARMv7 targets (#26161)

weidu-tpvision committed 87 days ago

Verified 21095477

[CUDA] Fix cuda 13 build (#26153)

tianleiwu committed 87 days ago

Verified 4de31529

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

apsonawane committed 87 days ago

Verified 42fcd71a

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

quic-calvnguy committed 87 days ago

Verified b160e8c6

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

quic-muchhsu committed 87 days ago

Verified b608f798

MultiheadAttention CUDA BF16 Support (#26083)

nenad1002 committed 87 days ago

Verified b7ae53f7

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

daijh committed 88 days ago

Verified f2f50ebc

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

kunal-vaishnavi committed 88 days ago

Verified 99e10e68

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

quic-ashwshan committed 88 days ago

Verified 8c5d2b58

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

xadupre committed 88 days ago

Verified 6650e07b

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

adrianlizarraga committed 88 days ago

Verified 88f2652d

[C#] Implement missing APIs (#26101)

yuslepukhin committed 88 days ago

Verified 35dcab50

Older