microsoft/onnxruntime

Pull Requests Commits

add missing atomic include (#9440)

jywu-msft committed 4 years ago

Verified 38738853

hipify tensor/gather_nd_impl.cu (#9392)

jeffdaily committed 4 years ago

Verified 52c53e39

hipify fast_divmod.h (#9400)

jeffdaily committed 4 years ago

Verified a2ba923a

hipify tensor/transpose.cc and tensor/transpose.h (#9397)

jeffdaily committed 4 years ago

Verified a8e2e8d7

Set cuda version to be None instead of an empty string (#9435)

baijumeswani committed 4 years ago

Verified 757bc667

Update ORTTraiing frontend codeowner (#9427)

SherlockNoMad committed 4 years ago

Verified e22920d9

Kernels for QLinearConv with symmetrically quantized filter (#9323)

yufenglee committed 4 years ago

Verified da3dd398

Make FusedAdam mathematically equivalent to Transformers AdamW (#9343)

baijumeswani committed 4 years ago

Verified 5da4e07d

fixes SDL Native Rules warning in Node.js binding CI (#9402)

fs-eire committed 4 years ago

Verified 5b65f1cb

Add support for DmlExecutionProvider for transformer profiler tool (#9380)

Jingqiao Fu committed 4 years ago

Verified f60e6030

Add Dev Guide to transformer optimizer (#9329)

wangyems committed 4 years ago

Verified 0824207c

Update win-ci-pipeline.yml

Changming Sun committed 4 years ago

6ecb990f

fix MSVC micro benchmark build warnings (#9373)

tracysh committed 4 years ago

Verified b130a7b7

Fix integer overflow for large step for Slice OP (#9376)

guoyu-wang committed 4 years ago

Verified 59dfab59

[js/web] remove webgl from default fallback list (#9374)

fs-eire committed 4 years ago

Verified 901c7de9

Exception when duplicated autograd.Function name detected (#9351)

pengwa committed 4 years ago

Verified f05c285a

[js/web] Support opset-13 for squeeze, unsqueeze, maxpool, pad, cast and clip (#9249)

hanbitmyths committed 4 years ago

Verified 74eaaad7

[ROCm] static re-hipify of CUDA EP to ROCm EP, now a shared provider (#8877)

jeffdaily committed 4 years ago

Verified c8789d30

Abjindal/merge eager with external custom ops (#8986)

ajindal1 committed 4 years ago

Verified 87e726d1

Abjindal/eager windows build (#9326)

ajindal1 committed 4 years ago

Verified 23700a15

work around ucx in rocm ci Dockerfile (#9360)

jeffdaily committed 4 years ago

Verified 3e879aab

Remove tensorflow, tf2onnx from the list of dependencies for the documentation (#9221)

xadupre committed 4 years ago

Verified 11f0081c

Refactor TrainingManager.forward (#9354)

xadupre committed 4 years ago

Verified 22e3f8bf

[DML EP] ConstantsOfShape - Empty Output and EinSum - Optional Parameter (#9361)

sumitsays committed 4 years ago

Verified 85155453

legacy_megatron-lm/deepspeed_ZERO1&2 FP16_Optimizer wrapper (#9184)

pengwa committed 4 years ago

Verified 5ee47e3f

fix to avoid quantizing attention with varied q,k,v sizes (#9357)

viboga committed 4 years ago

Verified 4771256b

Hooked up eager logging to ORT default logger. (#9340)

chandru-r committed 4 years ago

Verified ba0cca96

Add Conv and ConvTrans to ROCm EP (#9338)

groenenboomj committed 4 years ago

Verified 905fe365

Serizalize model only once to reduce backend preparation overhead (#8270)

IceTDrinker committed 4 years ago

Verified bccd09c6

Add Transpose, Reshape, Pow and LeakyRelu ops to DNNL execution provider (#9180)

georgen117 committed 4 years ago

Verified e8ba5145

Newer Older