microsoft/onnxruntime

Pull Requests Commits

Rename function to avoid problem of only one Alloc is overriden in derived class

cloudhan committed 3 years ago

499834ae

Switch to TensorDesc for datatype and shape aware allocation

cloudhan committed 3 years ago

c49ad561

Revert change over ortdevice.h

cloudhan committed 3 years ago

fb951b66

[core] Slightly extends allocator interface

cloudhan committed 3 years ago

feb3423d

Exclude EPs that aren't available on mobile to try and fix Xamarin build error on M1. (#11267)

skottmckay committed 3 years ago

Verified c5de493a

Consolidate utils::ToTensorProtoElementType, TypeToDataType, and data_types_internal::ToTensorDataType. (#9824)

edgchen1 committed 3 years ago

Verified 4854a093

Fix an SAL annotation error

Changming Sun committed 3 years ago

2cacd18d

Move gpt2 script to models\gpt2 sub-directory (#11256)

tianleiwu committed 3 years ago

Verified 1d96cbec

Model tests refactor (#11194)

chilo-ms committed 3 years ago

Verified cb46d791

Increase the timeout so the packaging pipeline stops failing.

skottmckay committed 3 years ago

af249943

[MinBuild] 132KB minimal build binary size reduction via dummy __cxa_demangle (#11071)

cloudhan committed 3 years ago

Verified 013306c9

Update QDQFinalCleanup transformer to also handle removing DQ/Q node pairs. (#11219)

edgchen1 committed 3 years ago

Verified 180b3f7c

Fix NNAPI EP error when handling external node adjacent to partition. (#11233)

edgchen1 committed 3 years ago

Verified e3ff4a6b

Support only one input in QLinearConcat (#11265)

zhanghuanrong committed 3 years ago

Verified 70d97bdf

remove deprecated quantize api (#11263)

yufenglee committed 3 years ago

Verified 2e6c2177

ppc64le: Optimizing the MlasMaximumPool() to use VSX instructions (#11216)

maxiwell committed 3 years ago

Verified acb555c4

auto mixed precision for t5 (#11252)

tianleiwu committed 3 years ago

Verified bab9b80f

[js] use NPM and yarn to upgrade package version (#11059)

fs-eire committed 3 years ago

Verified 5ee8e2e4

SizeInBytes Fix for Strided Tensor (#11224)

Lafi7e committed 3 years ago

Verified 06026fe8

Add option to specify onnxruntime repo URL in tools/android_custom_build/build_custom_android_package.py. (#11250)

edgchen1 committed 3 years ago

Verified 3dac6669

Fix find_package for benchmark

lukasberbuer committed 3 years ago

efb0928e

Scoped GIL release in run_with_iobinding (#11248)

yuslepukhin committed 3 years ago

Verified 98faaa7e

Improve onnx shape inference in quant tool (#11106)

yufenglee committed 3 years ago

Verified dec99657

fix build warnings (#11213)

pengwa committed 3 years ago

Verified 9765ef8b

[CUDA] Rollback TileMemcpy and TileBatchedMemcpy when Block Size is Small (#11187)

Lafi7e committed 3 years ago

Verified 0bad5b1b

One dnn v2.6 update (#11220)

georgen117 committed 3 years ago

Verified d9eeb483

Fixed compilation error for ARM architecture (#11223)

sumitsays committed 3 years ago

Verified 227bc726

MatMulScaleFusion - handling scale input (#11121)

ytaous committed 3 years ago

Verified bc296c70

use int storage (#11185)

mszhanyi committed 3 years ago

Verified 94032357

add AppendExecutionProvider_CUDA_V2 to the C++ api (#11153)

Ahmad Zakaria committed 3 years ago

Verified 63ff391b

Older