microsoft/onnxruntime

Pull Requests Commits

provide device holder

Cheng Tang committed 2 years ago

922f51e9

comment out initializing root_stream

jslhcl committed 2 years ago

4ea7d9b7

no change on execution_frame

jslhcl committed 2 years ago

7c1e7fce

Merge branch 'main' of https://github.com/microsoft/onnxruntime into leca/memPatternThreadSafe3

jslhcl committed 2 years ago

cde88213

Exposing new TRT build options (#15089)

gedoensmax committed 2 years ago

Verified fbe88fcc

Add compilation cache in 2 Linux CPU pipelines and refactor the Linux build step with cache (#15484)

mszhanyi committed 2 years ago

Verified 4e1f7581

Share more constant initializers (#15461)

pengwa committed 2 years ago

Verified bf32dbbd

Fix an indent error in build.py (#15497)

Changming Sun committed 2 years ago

Verified f297bbb8

[ROCm] Add hipBLASLt GEMM support to Tunable op. (#15351)

mindest committed 2 years ago

Verified 0fdd356a

SkipLayerNorm fusion with different input and output type (#15500)

hanbitmyths committed 2 years ago

Verified fda0aa14

Capture both ValueError and RuntimeError (#15503)

wschin committed 2 years ago

Verified d76cf374

Add support to use sequence as input ids in decoder inputs to Beam Search CUDA Op (#15232)

apsonawane committed 2 years ago

Verified 56ad6812

Disable XNNPack EP's tests in Windows CI pipeline (#15406)

Changming Sun committed 2 years ago

Verified 5bed8d02

test pipeline by using Stream directly

jslhcl committed 2 years ago

fbb8d586

softmax perf improvement pr2 - import softmax bw (#15199)

zhijxu-MS committed 2 years ago

Verified 05ec2233

disable BatchNormalizationGrad test (#15485)

mindest committed 2 years ago

Verified 67ac3610

Add a missing header to cuda_common.h (#15489)

Changming Sun committed 2 years ago

Verified a22dc65a

Optimize SCE loss compute (#15401)

pengwa committed 2 years ago

Verified 516c8e95

Remove codecov from requirements-dev.txt (#15487)

justinchuby committed 2 years ago

Verified 07b64d52

[DML EP] Use ORT node names in DML execution plans (#15411)

PatriceVignola committed 2 years ago

Verified fd7f0c3c

Fix bug in handling of variadics in function schema creation (#15409)

gramalingam committed 2 years ago

Verified e361e3f1

[build/npm] dump ORT_COMMON_FROM from validation (#15475)

fs-eire committed 2 years ago

Verified e1e88522

[build] fix default value of flag cmake_generator (#15471)

fs-eire committed 2 years ago

Verified 3875c824

[build] fix nuget linux in build_protoc_for_host() (#15472)

fs-eire committed 2 years ago

Verified 041a0e27

[TensorRT EP] support TensorRT 8.6-EA (#15299)

yf711 committed 2 years ago

Verified 8cd5f3ad

Move DML CI Pipeline to A10 (#15468)

nums11 committed 2 years ago

Verified e3086b2e

remove python test to pass pipeline

jslhcl committed 2 years ago

8440ee49

[ROCm] disable composable_kernel and kernel explorer for MIGraphX CI (#15479)

PeixuanZuo committed 2 years ago

Verified 00165540

[ROCm] add FP16 support for FusedConv Op (#15443)

PeixuanZuo committed 2 years ago

Verified d49a8de9

[ROCm] Add Tunable GroupNorm (#15298)

PeixuanZuo committed 2 years ago

Verified ce1eb6d6

Older