microsoft/onnxruntime

Pull Requests Commits

Minor fix for differently scoped cpu_ep usage (#15550)

pengwa committed 2 years ago

Verified d8dfda2e

Refactor FusedConv test (#15512)

PeixuanZuo committed 2 years ago

Verified 8bec6cd0

Use os.fspath on Path (#15530)

justinchuby committed 2 years ago

Verified 9d26f8f4

chilo-ms committed 2 years ago

5c5aaa81

chilo-ms committed 2 years ago

12e8613b

chilo-ms committed 2 years ago

4e93566e

chilo-ms committed 2 years ago

ca9dbfa3

Merge branch 'main' into chi/explicit_trt_profiles

chilo-ms committed 2 years ago

Verified 1661f309

Fix/Enhance convert_generation tool for SkipLayerNorm, op_block_list... (#15368)

zhanghuanrong committed 2 years ago

Verified a30b57da

chilo-ms committed 2 years ago

2beeb76a

chilo-ms committed 2 years ago

7109ce96

chilo-ms committed 2 years ago

93ef13ea

Bump ruff in CI (#15533)

justinchuby committed 2 years ago

Verified a36caba0

create a stream in DeviceStreamCollection for memory pattern (#15426)

jslhcl committed 2 years ago

Verified c2221d91

enable training tests for csharp bindings (#15513)

askhade committed 2 years ago

Verified 8fa65aba

Fix FusedConv for ROCm (#15460)

cloudhan committed 2 years ago

Verified 7ed3bfde

Force using fixed random seeds for flaky tests (#15515)

wschin committed 2 years ago

Verified ac6ceffb

chilo-ms committed 2 years ago

4c2c8395

[QNN EP] Fix pool and conv op tests (#15504)

adrianlizarraga committed 2 years ago

Verified 5ebe700a

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

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

Newer Older