microsoft/onnxruntime

Pull Requests Commits

Fix softmax export (#20057)

pengwa committed 1 year ago

Verified 1a0ba3f6

[QNN QDQ Quant] Utils to generate mixed-precision quant overrides (#20028)

adrianlizarraga committed 1 year ago

Verified 7d976cf7

Add Symbolic Shape Hint to Triton Codegen Config (#20056)

Lafi7e committed 1 year ago

Verified d30c81d2

Packed QKV and Rotary Embedding Support for sm<80 GQA (#20012)

aciddelgado committed 1 year ago

Verified 4a196d15

Fix issue that failed to load Conv node with external initializer (#20042)

HectorSVC committed 1 year ago

Verified f977be06

[JS/WebGPU] Use non-matmul implementation for ConvTranspose in channel-first case. (#20022)

satyajandhyala committed 1 year ago

Verified 5b64d7c3

[QDQ Quant] Support mixed-precision integer quantization via overrides (#19925)

adrianlizarraga committed 1 year ago

Verified cdc5d72b

Fix a bug in WASM's GEMM (#20023)

Changming Sun committed 1 year ago

Verified 3b4b99b9

Add ModelProto support for transformers optimize_model (#19990)

xiaoyu-work committed 1 year ago

Verified 71551dac

Make MS Debug engine SymInitialize() called as needed. (#20036)

yuslepukhin committed 1 year ago

Verified 3076b569

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

kunal-vaishnavi committed 1 year ago

Verified f9cddd2c

remove const cast for DLManagedTensor (#20015)

ajindal1 committed 1 year ago

Verified 7e84ba0e

Support model with multiple SCE loss nodes (#20016)

baijumeswani committed 1 year ago

Verified 2bc29244

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

kunal-vaishnavi committed 1 year ago

Verified 6238e9c0

Ort openvino npu 1.17 master (#19966)

sfatimar committed 1 year ago

Verified eab35c20

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

mszhanyi committed 1 year ago

Verified cd6d3aea

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

Changming Sun committed 1 year ago

Verified dafbef3a

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

tpboudreau committed 1 year ago

Verified 983fd839

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

mszhanyi committed 1 year ago

Verified 30a0d809

Increase MNIST test tolerance (#20000)

tianleiwu committed 1 year ago

Verified 06fe4f31

add random seed to layernorm tests (#19998)

prathikr committed 1 year ago

Verified 0b958bb4

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

mszhanyi committed 1 year ago

Verified 175f149b

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

justinchuby committed 1 year ago

Verified 0335ea9f

turn on neural_speed by default (#19627)

yufenglee committed 1 year ago

Verified 15219e2e

Support xcframework for mac catalyst builds. (#19534)

YUNQIUGUO committed 1 year ago

Verified 6b305f95

String Tensor SplitToSequence fix (#19942)

Craigacp committed 1 year ago

Verified 19ff4a6d

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

mtavenrath committed 1 year ago

Verified 0af5eacc

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

mszhanyi committed 1 year ago

Verified 8adbc093

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

zesongw committed 1 year ago

Verified 7e18cb4c

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

wangyems committed 1 year ago

Verified 6ff31e06

Newer Older