microsoft/onnxruntime

Pull Requests Commits

TRT 10 supports INT64

chilo-ms committed 2 years ago

3d3571ad

Remove early stopping from LLaMA end-to-end benchmarking (#20033)

kunal-vaishnavi committed 2 years ago

Verified f9cddd2c

remove const cast for DLManagedTensor (#20015)

ajindal1 committed 2 years ago

Verified 7e84ba0e

Support model with multiple SCE loss nodes (#20016)

baijumeswani committed 2 years ago

Verified 2bc29244

Add LLaMA end-to-end benchmarking (#19985)

kunal-vaishnavi committed 2 years ago

Verified 6238e9c0

Ort openvino npu 1.17 master (#19966)

sfatimar committed 2 years ago

Verified eab35c20

Refactor Python CUDA packaging pipeline to fix random hangs in building (#19989)

mszhanyi committed 2 years ago

Verified cd6d3aea

CMake: support reading dependency zip files from a local mirror (#20005)

Changming Sun committed 2 years ago

Verified dafbef3a

Recognize NaN operands in Min and Max ops (#19984)

tpboudreau committed 2 years ago

Verified 983fd839

Fix exception in Publish unit test results step (#20007)

mszhanyi committed 2 years ago

Verified 30a0d809

Increase MNIST test tolerance (#20000)

tianleiwu committed 2 years ago

Verified 06fe4f31

add random seed to layernorm tests (#19998)

prathikr committed 2 years ago

Verified 0b958bb4

Remove downloading deps in CUDA package test stage (#19993)

mszhanyi committed 2 years ago

Verified 175f149b

Use Java 11 to build project in the codeql pipeline (#19999)

justinchuby committed 2 years ago

Verified 0335ea9f

turn on neural_speed by default (#19627)

yufenglee committed 2 years ago

Verified 15219e2e

Support xcframework for mac catalyst builds. (#19534)

YUNQIUGUO committed 2 years ago

Verified 6b305f95

String Tensor SplitToSequence fix (#19942)

Craigacp committed 2 years ago

Verified 19ff4a6d

Fix broken Pooling CUDA NHWC Ops and ensure NCHW / NHWC parity. (#19889)

mtavenrath committed 2 years ago

Verified 0af5eacc

[Fix] Error Python Packaging Pipeline (Training CPU) (#19992)

mszhanyi committed 2 years ago

Verified 8adbc093

[WebNN EP] Support MatMul 1D (#19862)

zesongw committed 2 years ago

Verified 7e18cb4c

[MoE] Add TP and Mixtral MoE (#19945)

wangyems committed 2 years ago

Verified 6ff31e06

[ROCm] Remove MPI dependency and collectives to use NCCL (#19830)

mindest committed 2 years ago

Verified 3dfe4a5e

Add const cast for DLManagedTensor (#19982)

ajindal1 committed 2 years ago

Verified 6fe02068

[js/webgpu] fix maxpool / fp16 (#19981)

guschmue committed 2 years ago

Verified c45cff60

Adjust test tolerance (#19947)

tianleiwu committed 2 years ago

Verified 597e828a

Switch a portion of CI/packaging jobs to MacOS12 (#19908)

hariharans29 committed 2 years ago

Verified cd6ec50b

[NhwcTransformerTests] Fix linker error due to explicit template instantiation of ModelBuilder methods (#19980)

adrianlizarraga committed 2 years ago

Verified 18a7f34b

[js/webgpu] allow setting env.webgpu.adapter (#19940)

fs-eire committed 2 years ago

Verified 01c7aaf6

Exclude TRT provider in tests crashed in A100 (#19972)

tianleiwu committed 2 years ago

Verified 8293aa15

Fix Training CPU docker image name to avoid unnecessary rebuilding (#19973)

mszhanyi committed 2 years ago

Verified d4c8bc35

Older