microsoft/onnxruntime

Pull Requests Commits

Fix training python packaging pipeline (#6613)

Changming Sun committed 5 years ago

Verified e70344e6

Update a few WinML model test filters for DML

jstoecker committed 5 years ago

Verified 1c727742

Support disabling training kernels as part of a reduced build (#6557)

Cian Hayes committed 5 years ago

Verified 8f14b8bd

Add engine decryption in TensorRT EP (#6612)

stevenlix committed 5 years ago

Verified e9d03983

Update CUDA build configs (#6598)

Changming Sun committed 5 years ago

Verified 0b89f931

Support double for operators ReduceMax, ReduceMin (#6265)

xadupre committed 5 years ago

Verified d3a2c8c1

enable omp for debug build

RandyShuai committed 5 years ago

ff063309

deprecate omp in ci

RandyShuai committed 5 years ago

6c5f50d0

Quantize model with QDQ format (#6541)

yufenglee committed 5 years ago

Verified 56e4e47f

Make kernel hash stable in type reduced build (#6603)

skottmckay committed 5 years ago

Verified c02ae61c

Fix layer_norm.cc on x86 (#6556)

Cian Hayes committed 5 years ago

Verified 16eed68a

Don't update the excluded ops/types unless args.update is true. Updating the exclusion info triggers rebuilding of all kernels using type reduction. (#6604)

skottmckay committed 5 years ago

Verified 13d7db9a

Fix bug with ORT format serialization of tensor attributes. (#6602)

skottmckay committed 5 years ago

Verified 0b1e21c6

[Mult-GPU inferencing] Add new API to get/set device id. Set correct device id in cuda allocator. (#6592)

Pranav Sharma committed 5 years ago

Verified 67ef6b1a

Fix TensorRT unnecessary file cache operations (#6601)

nietras committed 5 years ago

Verified 1dd920fa

Reduce CastMLFloat16ThroughFloat size (Scott's suggested changes), fix unused function warning. (#6597)

edgchen1 committed 5 years ago

Verified 19c130f5

Fix some coding conventions issues (#6583)

skottmckay committed 5 years ago

Verified 190b90a6

Generate error when an explicit stream argument is not provided in the <<<...>>> kernel launch syntax (#6599)

weixingzhang committed 5 years ago

Verified c86c21e0

Enable more ROCM ops that are sharing CUDA code. Some are needed for Turing NLG models.

jessebenson committed 5 years ago

d18aa45b

Fix build.gradle so it always targets Java 8 class files.

Craigacp committed 5 years ago

dbe31361

Delay load dxcore in winml model tests

Ryan Lai committed 5 years ago

b57a7f4d

Fix build failure when building with --build_wheel on Windows

georgen117 committed 5 years ago

b50b0a89

Remove docs that have been migrated to https://onnxruntime.ai/docs (#6225)

natke committed 5 years ago

Verified af9dfa7a

Fix updated Doxygen errors. (#6588)

yuslepukhin committed 5 years ago

Verified dda5a620

ort_test_utils: skip creating input if it is an initializer (#6544)

jcwchen committed 5 years ago

Verified 115e16b3

Remove condition from ORT_RETURN_IF[_NOT] macro output. (#6563)

skottmckay committed 5 years ago

Verified ccfd9029

Update GPU packaging pipelines to cuda11 and fix the other build break issues (#6585)

Changming Sun committed 5 years ago

Verified b5bd14fc

Support no bias in layernorm and skiplayernorm op (#6554)

wangyems committed 5 years ago

Verified 82229c8e

Support to allow user to specify compute stream per session (#3723)

weixingzhang committed 5 years ago

Verified 299ace07

OpenVino add build_shared_lib flag in the build command (#6560)

sfatimar committed 5 years ago

Verified 973c3917

Newer Older