microsoft/onnxruntime

Pull Requests Commits

Enable C# GPU tests in Windows GPU CI pipeline (#8142)

Changming Sun committed 4 years ago

Verified a41d0db4

Enable TRT provider option configuration for C# (updated version) (#7808)

chilo-ms committed 4 years ago

Verified 91075255

Fix Python Cuda loading issues (#7939)

RyanUnderhill committed 4 years ago

Verified 49938cce

Use std::make_reverse_iterator directly

Changming Sun committed 4 years ago

378a9859

Fetching frontier tensors to frontend for ORTModule (#8086)

ashbhandare committed 4 years ago

Verified 00e44861

Enable signed int8 data type for activations in static quantization (#7029)

SilvanK4t1qbit committed 4 years ago

Verified eb36258d

Disable InitProvidersSharedLibrary when training is enabled. (#8132)

RyanUnderhill committed 4 years ago

Verified e083d207

[Java] Fixing the creation of OnnxTensors from scalars, adding tests (#8023)

Craigacp committed 4 years ago

Verified 7ed9f5fc

Adding optional ops in contrib ops (#7946)

neginraoof committed 4 years ago

Verified 80b7b134

Ortmodule override torch.manual_seed() (#8131)

SherlockNoMad committed 4 years ago

Verified 59e33604

Fuse attention node even in case of different Q,K hidden dimensions (#8106)

viboga committed 4 years ago

Verified b478086b

Update logic in props.xml to account for shared provider library changes (#8138)

hariharans29 committed 4 years ago

Verified 4fd7efcf

Update run_dockerbuild.sh: set default python version based on OS version (#8136)

Changming Sun committed 4 years ago

Verified f000dfdd

Chang how numpy version is handled. (#8130)

Changming Sun committed 4 years ago

Verified 1fa69866

[js] fixing broadcast issues in pack mode (#8090)

Tixxx committed 4 years ago

Verified db88f305

MLAS: enable SSE 4.1 path for x86 build (#8127)

tracysh committed 4 years ago

Verified cbdd59da

User dynamic axes in one step beam search output (#8092)

xi-liu-ds committed 4 years ago

Verified 45ce2399

Add int64 as a required type to ConstantOfShape as it's used by the pytorch converter for Pad. (#8128)

skottmckay committed 4 years ago

Verified cccd61e3

[Mobile package] Update required operator config with additional ops for newer version of Wav2Vec 2. (#8123)

edgchen1 committed 4 years ago

Verified b1e21312

Col2im optimization by eliminating integer multiplications:

Evgenii Indenbom committed 4 years ago

664e548e

Delete some unused code in run_dockerbuild.sh and Enable Nuget CUDA tests (#8089)

Changming Sun committed 4 years ago

Verified 6e2b064a

[Android] Output error message to android log instead of stderr (#8114)

guoyu-wang committed 4 years ago

Verified f6292d9b

Fix 32bit Android java API crash (#8122)

guoyu-wang committed 4 years ago

Verified 9003df5d

initialize generated_value_names with graph input (#8085)

yufenglee committed 4 years ago

Verified 4bb0e29d

Implement WINRT_IMPL_LoadLibraryW to avoid calling LoadLibraryW directly (#8065)

Ryan Lai committed 4 years ago

Verified 839f69d2

Update migraphx to rocm4.2 (#7994)

scxiao committed 4 years ago

Verified e7d7fa8f

Fix a compile error in InferenceTest.cs (#8119)

Changming Sun committed 4 years ago

Verified 5809890b

remove debug.keystore from repository due to a credential issue report (#8113)

hanbitmyths committed 4 years ago

Verified 8cacb269

Add TRT 7.1 Pipeline (#8073)

chilo-ms committed 4 years ago

Verified 27d1784d

Added support for ReduceMean on DNNL EP for CPU and GPU (#7902)

chethanpk committed 4 years ago

Verified 3cd06cb3

Older