microsoft/onnxruntime

Pull Requests Commits

Add comprehensive tests for GQA unfused attention (PR #28198)

Copilot committed 60 days ago

Verified 10215a18

tianleiwu committed 61 days ago

6354c525

Add C++ tests for GQA unfused attention with large head_size

tianleiwu committed 61 days ago

778ee069

fix: address review summary feedback - SafeInt, logging, tests, v_head_size

tianleiwu committed 61 days ago

b0f71ac0

fix: address review feedback - SafeInt AlignTo, y_bnsh H_v, ORT_ENFORCE

tianleiwu committed 61 days ago

eeec5123

review feedback

tianleiwu committed 61 days ago

4fa20eff

GQA unfused attention with FP32 QK accumulation (fixes #28195)

tianleiwu committed 62 days ago

99aa8b22

Fix string tensor deserialization in ORT format models (#28133)

Copilot committed 62 days ago

Verified c5e6bd8d

Add MatMul double implementation, inspect optimizers add test coverage (#28145)

yuslepukhin committed 62 days ago

Verified 3729b614

Test Driver update (#28190)

eserscor committed 62 days ago

Verified 73378141

Apply model size limit before converting size_t input to int32_t (#28112)

skottmckay committed 62 days ago

Verified 564cabfc

Enable API doc generation builds in PR checks (#28124)

edgchen1 committed 63 days ago

Verified a3d6452b

Implement GetAvaiableResource() callback (#28103)

yuslepukhin committed 63 days ago

Verified b20f8fce

[webgpu] support arbitrary input_channel size for im2col (#27038)

wenqinI committed 63 days ago

Verified 43d7cf4d

Fix lack of auth on python packaging (#28118)

eserscor committed 63 days ago

Verified dce77a36

Fix RemoveDuplicateCastTransformer incorrectly eliminating lossy Cast chains targeting bool (#28102)

Copilot committed 64 days ago

Verified 19052493

Add seqlens_k bounds validation in GroupQueryAttention to prevent GEMM OOB (#28031)

vraspar committed 64 days ago

Verified 7c56fa83

CoreML: Add support for Pad with 'reflect' for ML Program (#28073)

skottmckay committed 64 days ago

Verified fb13eb3e

Fix doxygen issues (#28122)

yuslepukhin committed 64 days ago

Verified 2adf3292

ICM fixes (1/n) (#27906)

hariharans29 committed 64 days ago

Verified 8ca8dd3c

[NvTensorRTRTX EP] Implement GetHardwareDeviceIncompatibilityDetails with driver and compute capability checks (#27577)

umangb-09 committed 64 days ago

Verified 7a33d4b2

Fix packaging pipeline TSA Upload warnings (#28127)

skottmckay committed 65 days ago

Verified 2070b28c

Add INT8, INT16, and UINT8 type support for CUDA TopK operator (#27862)

elwhyjay committed 65 days ago

Verified f0180666

Use cuBLAS status APIs for error strings (#27946)

tianleiwu committed 67 days ago

Verified d354e56f

WebGPU plugin EP pipeline updates (#28121)

edgchen1 committed 67 days ago

Verified a7790770

Address build failure on ARM (#28120)

yuslepukhin committed 67 days ago

Verified 9e1621df

Disable QDQ related transformers for WebNN EP (#28020)

miaobin committed 67 days ago

Verified e3063083

Use weights_only for torch.load checkpoints (#28097)

tianleiwu committed 67 days ago

Verified c5bc801f

Use SafeInt for size arithmetic in CPU tensor operators to prevent overflow (#28060)

tianleiwu committed 68 days ago

Verified a208df8a

Update worker thread pool to use time based wait. (#27916)

sushraja-msft committed 68 days ago

Verified 3f74b3cf

Older