microsoft/onnxruntime

Pull Requests Commits

try removing checkout and disabling tsa instead

edgchen1 committed 22 days ago

Verified 49779ec1

Fill opset gap for RandomNormal, RandomNormalLike, RandomUniform, RandomUniformLike CUDA ops with BFloat16 support (#27759)

Copilot committed 22 days ago

Verified 0139499a

Fixes for unicode path handling (#28390)

jnagi-intel committed 22 days ago

Verified 9b3fb24a

Add float zero point support for 2-bit LUT GEMM in MatMulNBits (#28354)

vraspar committed 22 days ago

Verified e8ae6ce9

Bound total output allocation size in Tile kernel (#28070)

GopalakrishnanN committed 22 days ago

Verified a363e895

Fix OOB reads in SoftmaxCrossEntropyLoss via label bounds validation (#28004)

vraspar committed 22 days ago

Verified ad11f7cf

Use weights_only=True for remaining torch.load() calls (#28421)

adrianlizarraga committed 22 days ago

Verified 4e377a13

Add LabelEncoder_4 fast path for session creation (#27526)

SamuelLess committed 22 days ago

Verified 73d54165

Add RVV (RISC-V Vector Extension) optimized convolution and pooling kernels for the NCHWc blocked format in MLAS (#28411)

velonica0 committed 23 days ago

Verified 0e721889

Address Resize kernel shortcomings (#28402)

yuslepukhin committed 23 days ago

Verified 469bc9f6

[webgpu] Optimize LinearAttention Op with subgroup (#28412)

daijh committed 23 days ago

Verified a8260cd6

Fill CUDA EP opset gap for Sin and Cos operators (7→22) (#27756)

Copilot committed 23 days ago

Verified 480aad09

Use symbol file to export symbols in onnxruntime library on AIX platform as well (#28016)

ayappanec committed 23 days ago

Verified 239ef9e3

[WebGPU] Correct MatMul bias input indexing in `MatMul::ComputeInternal` (#28475)

Copilot committed 23 days ago

Verified 72bce1df

fix: skip DQ->MatMulNBits fusion when weight/scale initializer is shared (#28326)

Rishi-Dave committed 23 days ago

Verified 183e7a94

[WebGPU] Fix SkipSimplifiedLayerNormalization bias (#28427)

xenova committed 23 days ago

Verified fafe5644

Add CPU QMoE 2-bit support and LUT GEMM fast path (#28185)

tianleiwu committed 23 days ago

Verified dac5a143

[ Cuda] ConvTranspose-22 (#27710)

tianleiwu committed 23 days ago

Verified a12b611e

Fix sigmoid transformation in TreeEnsembleClassifier for all-positive weights with LOGISTIC post_transform (#27536)

Copilot committed 23 days ago

Verified 756a5235

Enable QuickGeluFusion on WebGPU EP and fix fp16 shader (#28410)

xiaofeihan1 committed 23 days ago

Verified a49da870

Fix runtime-unresolvable type annotations in Session and InferenceSession (#27802)

Rishi-Dave committed 23 days ago

Verified 8d24011e

Add DiT attention fusion for F5-TTS and diffusion transformer models (#27999)

Rishi-Dave committed 23 days ago

Verified ece097dc

Add CPU Cast op support for Float8E8M0 (#28435)

tianleiwu committed 23 days ago

Verified 8cfb63e0

[CUDA] PagedAttention: use exact max_query_len on FA path (#28409)

elwhyjay committed 23 days ago

Verified d82a06df

[GQA] Make present_key/present_value outputs optional and add Gemma4 support (#28242)

apsonawane committed 23 days ago

Verified 83e402e0

Validate NodeArg creation (#28429)

yuslepukhin committed 23 days ago

Verified 51c546b0

Fix CUDA plugin EP test pipeline (#28453)

tianleiwu committed 23 days ago

Verified c1a61b44

Address tokenizer shortcomings (#28428)

yuslepukhin committed 23 days ago

Verified 67d9400d

Fix Clang + libc++ compilation and portability issues (#28049)

mustjab committed 24 days ago

Verified c6d9a020

Handle empty initializers gracefully in optimizer passes (#27976)

GopalakrishnanN committed 24 days ago

Verified 1b64fba3

Older