microsoft/onnxruntime

Pull Requests Commits

Fix C header to stop people accidentally copying the OrtApi by value (#12297)

RyanUnderhill committed 4 years ago

Verified 3e014a5e

Remove aten::binary_cross_entropy_with_logits from ATen Fallback (#12301)

Lafi7e committed 4 years ago

Verified c40f73ae

Eliminate memory allocations per recent profiling (#12225)

yuslepukhin committed 4 years ago

Verified 3bf614fd

Bump terser from 5.7.0 to 5.14.2 in /js/common (#12248)

dependabot[bot] committed 4 years ago

Verified 972bb967

On device training CI pipeline (#11987)

baijumeswani committed 4 years ago

Verified ddb45e91

Apply project formatting rules to ort_aten.cpp (#12294)

jamill committed 4 years ago

Verified 8d0e86de

[CUDA] Add Strided Tensor Support for Expand->GatherElements for Training (#11976)

Lafi7e committed 4 years ago

Verified 0fa3aeb6

CPU AdamW implementation (#11978)

pengwa committed 4 years ago

Verified 75bda9f2

Initialize generated tensor data in onnxruntime_perf_test. (#12275)

edgchen1 committed 4 years ago

Verified 564dc323

support gpt2 model with greedy search (#12068)

gh-yewang committed 4 years ago

Verified 89ac61f4

Minor fixes (#12276)

edgchen1 committed 4 years ago

Verified cb351388

Bump ort version number (#11948)

RandySheriffH committed 4 years ago

Verified 0264a9c2

[NNAPI EP] Add some support for MatMul with batch inputs (#12261)

edgchen1 committed 4 years ago

Verified a71f7d33

OrtModule fix pytorch version comparison (#12280)

juanpaez22 committed 4 years ago

Verified 4f57da78

Fix memory consumption discrepancy (#12266)

pengwa committed 4 years ago

Verified feabafe5

Remove two lines in the Dockerfile for Github Codespace (#12278)

Changming Sun committed 4 years ago

Verified f2533d30

Merge pull request #12056 from microsoft/bmeswani/merge-training_dev/on_device_poc

askhade committed 4 years ago

Verified ceb76429

fix bug: output of Concat is quantized twice in qdq format (#12254)

yufenglee committed 4 years ago

Verified 7194ec18

clean up calibration model (#12255)

yufenglee committed 4 years ago

Verified a18b0805

Modify generator for eager to use all inputs for determining promote type. (#12268)

WilBrady committed 4 years ago

Verified 45c0be8a

Bump terser from 5.10.0 to 5.14.2 in /js/web (#12253)

dependabot[bot] committed 4 years ago

Verified 30ac6e87

[CoreML EP] Remove batch=1 restriction in depthtospace op support (#12258)

YUNQIUGUO committed 4 years ago

Verified eb3b49b6

Add dev container / codespace configuration (#12256)

jamill committed 4 years ago

Verified 108b860d

[ROCm] Enable int8 for MatMulInteger Op (#11776)

xinyazhang committed 4 years ago

Verified 03dfcb0e

Make GetTrainingApi as a part of the OrtApis, add Training API documentation and address other pull request review comments

baijumeswani committed 4 years ago

cbf08c7a

Use unregister_custom_op_symbolic to unregister torch symbolics (#12146)

justinchuby committed 4 years ago

Verified 3d2bcb33

Update supported ops md for NNAPI/CoreML EP (#12245)

YUNQIUGUO committed 4 years ago

Verified 49661859

Implementing aten::gt.Scalar_out and aten::lt.Scalar_out (#12181)

LironKesem committed 4 years ago

Verified 7dc45bc3

Fix: Test coverage is undercounting and profiling errors (#12260)

mszhanyi committed 4 years ago

Verified 007ef427

Fix a bug in beam search custom attention mask allocation (#12240)

gh-yewang committed 4 years ago

Verified 5066ef11

Newer Older