microsoft/onnxruntime

Pull Requests Commits

Remove eigen device from thread pool

Changming Sun committed 5 years ago

edd5855f

fix python package linux gpu failure (#3786)

jywu-msft committed 5 years ago

Verified dcb1a215

Apply onnx-tensorrt bug fixes (#3785)

stevenlix committed 5 years ago

Verified 99ec93ea

Update documentation + Update mlas threading lib to use the new TrySimpleParallelFor. (#3779)

Pranav Sharma committed 5 years ago

Verified e42e0d47

disable cublasHgemm for training (#3769)

pengwa committed 5 years ago

Verified 29234458

collect the last few iteration latency for throuput calculation (#3766)

pengwa committed 5 years ago

Verified 98b97be6

avoid using LocalFree on FormatMessageA buffer (#3772)

Ori Levari committed 5 years ago

Verified ad63e259

Implement Pow(12) for cpu and cuda (#3727)

yuslepukhin committed 5 years ago

Verified f68a326b

Remove usage of openmp in reverse seq impl. (#3754)

Pranav Sharma committed 5 years ago

Verified 027a3649

Revert softmax kernel implementation (#3753)

Changming Sun committed 5 years ago

Verified 62c730a8

sync threads before calling next cub function (#3758)

RandySheriffH committed 5 years ago

Verified 86eaa71e

Liqun/e2e transformer test (#3540)

liqunfu committed 5 years ago

Verified af398819

Use cublasHgemm "back" for fp16 computation with Volta GPU (#3765)

pengwa committed 5 years ago

Verified 177c1357

Add Threadpool::TrySimpleParallelFor (#3759)

skottmckay committed 5 years ago

Verified 3421ec11

Add SoftmaxCrossEntropyLoss to mixed-precision-transformer. (#3760)

codemzs committed 5 years ago

Verified b9a5ed1f

Fix 'Install ONNX' CI failure (#3761)

skottmckay committed 5 years ago

Verified 9f727523

Refine GatherND CPU/CUDA Kernels & Add UTs (#3688)

pengwa committed 5 years ago

Verified 0531accc

Add Distributed Checkpointing support (#3639)

ashbhandare committed 5 years ago

Verified 58f53966

Properly creating arguments to pass to setup.py (#3744)

davidbrownellWork committed 5 years ago

Verified 7296e06d

fix warning treated as error due to ignoring return status (#3739)

suffiank committed 5 years ago

Verified ea0e2d1d

Limit the number of models run on OpenVINO (#3742)

suryasidd committed 5 years ago

Verified e529464a

Fix the test coverage pipeline (#3710)

Changming Sun committed 5 years ago

Verified 7ff06056

Fix evaluation issues (#3538)

Tixxx committed 5 years ago

Verified 0638565f

Fix flaky test and avoid divide by zero in SoftmaxCrossEntropyLoss-CPU. (#3734)

codemzs committed 5 years ago

Verified 939589c2

Fix a perf regression by providing a better estimate for the cost in LSTM's TryParallelFor call.

Pranav Sharma committed 5 years ago

bad90d7a

iOS cross build on MacOS (#3699)

guoyu-wang committed 5 years ago

Verified 12d7c2f6

Handle dim with value of zero in ConvTranspose (#3728)

skottmckay committed 5 years ago

Verified 29c12c0f

Merge pull request #3708 from microsoft/jeffbloo/MergeDmlDev

jeffbloo committed 5 years ago

Verified 9a4d1c77

Enable telemetry on windows zip packages (#3738)

smk2007 committed 5 years ago

Verified f1a948fd

add downlevel test artifact to windowsai-nuget build (#3711)

Ori Levari committed 5 years ago

Verified 78fde2c4

Newer Older