microsoft/onnxruntime

Pull Requests Commits

draft parameters bindings

AdamLouly committed 3 years ago

d82e10fb

fix reshape fusion error in numpy 1.24 (#14231)

tianleiwu committed 3 years ago

Verified 3b79b8eb

[DML EP] Fix FusedMatMul crash when batch > 1 (#14225)

PatriceVignola committed 3 years ago

Verified 52fc1ff2

Multi-tensor SGDOptimizer (on device training) (#14083)

pengwa committed 3 years ago

Verified a4180d79

Use json format to save and load partition config (#14169)

RandySheriffH committed 3 years ago

Verified ecd5ce0b

Improve TopP sampling (#14192)

yufenglee committed 3 years ago

Verified 7a9a6bce

Create dedicated build for training api (#14136)

askhade committed 3 years ago

Verified d92c663f

Bump json5 from 2.2.0 to 2.2.3 in /js/web (#14110)

dependabot[bot] committed 3 years ago

Verified 3a39736a

Improve rpb cuda kernel (#14195)

wangyems committed 3 years ago

Verified 342a5bf2

Tune test case for hybrid cpu (#14204)

RandySheriffH committed 3 years ago

Verified d152452d

Supporting Intel AMX instructions in quantized GEMM (#14042)

chenfucn committed 3 years ago

Verified 90142899

rename CrossAttention to MultiHeadAttention (#14201)

wangyems committed 3 years ago

Verified a01bf8db

pin ort-ext to 81e7799c69044c745239202085eb0a98f102937b (#14044)

guschmue committed 3 years ago

Verified 6b8c72cf

DML EP Register Identity-16 (#14053)

nums11 committed 3 years ago

Verified f4ea781b

Hot fix for prefast failure to unblock python package pipeline (#14206)

tianleiwu committed 3 years ago

Verified 05e26f30

Use a local copy of murmurhash3 in TensorRT shared library (#14207)

adrianlizarraga committed 3 years ago

Verified 3d8b596c

Add protected destructor to Provider structure (#14152)

RyanUnderhill committed 3 years ago

Verified da57c0a7

Add catch-all exception handler to API_IMPL_END (#14194)

RyanUnderhill committed 3 years ago

Verified f8117b6f

[MIGraphX] update the MIGraphX version used in ORT to rocm-5.4.0 (#14184)

PeixuanZuo committed 3 years ago

Verified 33367fa2

make WITHCACHE as an option in MacOS workflow (#14188)

mszhanyi committed 3 years ago

Verified 6463f438

update convert_generation for Attention op change (#14191)

tianleiwu committed 3 years ago

Verified 7e751ac6

[DML EP] Fix unconnected node removal logic (#14193)

PatriceVignola committed 3 years ago

Verified c151afec

[DML EP] Update DML_FEATURE_LEVEL 5.0 (#14172)

sumitsays committed 3 years ago

Verified 906f578b

to work with onnx 1.13 rc, implement ver 18 reduce and optioanl ops, … (#13765)

liqunfu committed 3 years ago

Verified 1be36913

Replace distutils by setuptools to import build_ext (#14108)

xadupre committed 3 years ago

Verified 79dc3960

[DML EP] Remove unconnected nodes from the graph (#14155)

PatriceVignola committed 3 years ago

Verified 64541a58

activate past_present_share_buffer for sampling node (#14166)

zhanghuanrong committed 3 years ago

Verified 74fe45bf

Add batched and strided batched gemm as TunableOp (#13841)

cloudhan committed 3 years ago

Verified be879c11

relational attention bias cuda op (#14149)

wangyems committed 3 years ago

Verified 5eac2c1f

Ignore more build directories and clangd files (#14154)

cloudhan committed 3 years ago

Verified 8e216301

Older