microsoft/onnxruntime

Pull Requests Commits

Improve symbolic shape inference in transformers tools (#12217)

tianleiwu committed 3 years ago

Verified 972e5e73

Eager mode - argmax_out: set output tensor (#12233)

jamill committed 3 years ago

Verified 975bb56e

merge cheng's change

Lei Cao committed 3 years ago

c9422e09

fix linux build break for PE

Lei Cao committed 3 years ago

2d183557

Fix GH issue 12208 (#12224)

yuslepukhin committed 3 years ago

Verified 555e8898

Simplify get_docker_image.py (#12166)

Changming Sun committed 3 years ago

Verified 2cb64292

Exclude implicit inputs from dump of encoder feeds in beam search (#12222)

tianleiwu committed 3 years ago

Verified 0c319d6e

[TVM EP][CI] Added TVMso EP testing into CI (#12188)

KJlaccHoeUM9l committed 3 years ago

Verified 66978c7e

Add eager mode support for mm.out (matrix multiplication). (#12214)

WilBrady committed 3 years ago

Verified 4235ebc1

[ROCM] Navi21 fixes pr (#11368)

micmelesse committed 3 years ago

Verified bb5bd085

remvoe useless code

chentaMS committed 3 years ago

db244d88

[CUDA] Split/Concat Kernel Optimization (#12175)

Lafi7e committed 3 years ago

Verified 173bcdbc

[js/web] use windowed Chrome for perf mode (#12157)

fs-eire committed 3 years ago

Verified ced7c2de

Add --disable_shape_inference option to optimizer.py (#12215)

tianleiwu committed 3 years ago

Verified b81b6526

Fix bug where onnxruntime_USE_NCCL flag would default to ON (#12195)

seanmurr1 committed 3 years ago

Verified 93229949

Merge branch 'PE' of https://github.com/microsoft/onnxruntime into PE

chentaMS committed 3 years ago

80d9ffb4

disable para plan test temporarily

RandySheriffH committed 3 years ago

24da06d5

remove parallel execution plan

chentaMS committed 3 years ago

a2c92042

remove identity in transformers model graph fusion (#12194)

tianleiwu committed 3 years ago

Verified 17b84c78

Add file mapping for windows platform. (#12183)

caoting-dotcom committed 3 years ago

Verified 4d38b84e

remove wrong placed libs (#12201)

leqiao-1 committed 3 years ago

Verified 09af4a7f

[TVM EP][CI] Integrate TVM EP into ORT public CI on Windows (#12161)

KJlaccHoeUM9l committed 3 years ago

Verified d31db1aa

fix cu file breaks

chentaMS committed 3 years ago

3dc15c95

Merge branch 'PE' into chenta/fix_training_build

chentaMS committed 3 years ago

ca27d42d

Fix line spacing/break issue, extend existing tests (#12191)

msftlincoln committed 3 years ago

Verified 52095fb0

refactor the cudnn call and cublas call

chentaMS committed 3 years ago

107d7aa4

OnnxRuntime Eager: Implement log_softmax with ONNX Ops (#12190)

msftlincoln committed 3 years ago

Verified a2dc6d32

enable only_execute_to_fetch and fix some training tests

chentaMS committed 3 years ago

d9a90251

bitwise_and ONNX support (#12189)

msftlincoln committed 3 years ago

Verified 9bca8405

Simple eager training models (#12180)

WilBrady committed 3 years ago

Verified 89bf6c9b

Newer Older