microsoft/onnxruntime

Pull Requests Commits

[ORTModule] Adjust Attention Patterns for Efficient Attention ATen Fallback (#18471)

Lafi7e committed 2 years ago

Verified 3bc9efc7

[QDQ Optimizer] Fix logic that drops Q/DQ ops from QDQ split node groups (#18394)

adrianlizarraga committed 2 years ago

Verified 7c573054

SDXL Latent Consistency Model (LCM) optimization (#18526)

tianleiwu committed 2 years ago

Verified 62da3b1c

[js/web] use Chrome in CI for npm tests (#18522)

fs-eire committed 2 years ago

Verified d455b0f8

[js/webgpu] enable f16 for concat (#18528)

qjia7 committed 2 years ago

Verified ac8598a8

Make TensorShapeVector to use InlinedVector<Int64_t> to reduce on template instantiations (#18519)

yuslepukhin committed 2 years ago

Verified 81a763a9

Training packaging pipeline for cuda12 (#18524)

ajindal1 committed 2 years ago

Verified 680a526e

Hide NPU Adapter selection behind macro (#18515)

smk2007 committed 2 years ago

Verified 2a016225

Add missing flags DISABLE_FLOAT8_TYPES in GemmFloat8 custom operator for CUDA < 11.8 (#18162)

xadupre committed 2 years ago

Verified 29a409ac

fix past-kv in general LLM exporter (#18529)

wejoncy committed 2 years ago

Verified a608c002

[js/web] unify resolve rules for "Clip" (#18527)

fs-eire committed 2 years ago

Verified c7fd9303

[js/webgpu] Optimize broadcast binary. (#18185)

qjia7 committed 2 years ago

Verified abdf8b7c

Create edges with arg positons correctly accounting for non-existing args (#18462)

yuslepukhin committed 2 years ago

Verified cc542024

[js] optimize eslint config (#18460)

fs-eire committed 2 years ago

Verified 247ce218

Remove setup_env_azure.bat (#18482)

jchen351 committed 2 years ago

Verified 1dd9bf53

Bfloat16 support for MatMulBnb4, Training support bitsandbytes>=0.41.2 (#18484)

jambayk committed 2 years ago

Verified 1af06815

Create a new Python Package pipeline for CUDA 12 (#18348)

jchen351 committed 2 years ago

Verified d97fc182

Tiny change to trigger the update of DORT's CI image (#18507)

wschin committed 2 years ago

Verified 3bcc137e

Update setup.py: replace libcudart.so.12.0 with libcudart.so.12 (#18501)

Changming Sun committed 2 years ago

Verified dc9ab4f8

Add fusion patterns for conformer-transducer model (#18461)

apsonawane committed 2 years ago

Verified 97cc40d7

Move up members in Lite Custom Op hierarchy for possible memleaks. (#18478)

RandySheriffH committed 2 years ago

Verified 53917a33

Update web-ci.yml: remove depth=1 (#18500)

Changming Sun committed 2 years ago

Verified 9364c051

[js] update a few packages (#18499)

fs-eire committed 2 years ago

Verified 34c54244

Removed all the deprecated python training code and related tests and utils (#18333)

askhade committed 2 years ago

Verified 02333293

[CoreML] Adapt to `MLMultiArray.dataPointer` deprecation (#17726)

NickLucche committed 2 years ago

Verified cbb85b48

Update NDK version to 26.1.10909125 (#18493)

Changming Sun committed 2 years ago

Verified 41f9379f

[js/web] JSEP Attention & MultiHeadAttention (#17742)

dakenf committed 2 years ago

Verified fac3e33d

[WebNN Ep] Slice's axes and steps inputs should be constant initializers (#18427)

Honry committed 2 years ago

Verified a5537f2f

rope support 4D input tensor (#18454)

kailums committed 2 years ago

Verified 1a294609

Always run emsdk_env.sh before build.py, even when ccache is disabled (#18477)

Changming Sun committed 2 years ago

Verified 5eb5056c

Newer Older