microsoft/onnxruntime

Pull Requests Commits

Update build docker image cache cleanup (#6048)

edgchen1 committed 5 years ago

Verified b348538c

Update profiler tool to support gpt2 and longformer models (#6011)

tianleiwu committed 5 years ago

Verified 51fbe87b

Remove python 3.8 Windows GPU build from python packaging pipeline (#6054)

Changming Sun committed 5 years ago

Verified 925879a8

fix windows cuda support for python 3.8 + (#6046)

jywu-msft committed 5 years ago

Verified 020efc90

Improve checkpointing for Zero stage 1 (#5478)

ashbhandare committed 5 years ago

Verified 7cebf76a

Update api_summary.rst (#6038)

hariharans29 committed 5 years ago

Verified a046ef13

Bump highlight.js from 10.2.1 to 10.4.1 in /nodejs

dependabot[bot] committed 5 years ago

d5e8c48e

Clean up builds (#6015)

edgchen1 committed 5 years ago

Verified d8139814

add missing iclosable interface (#6036)

smk2007 committed 5 years ago

Verified 00f43a3a

Use __launch_bounds__ workaround, rather than limiting threads to 256 on AMD.

jessebenson committed 5 years ago

14f6eb14

Re-enable Lamb unit tests for AMD

jessebenson committed 5 years ago

98ea7372

Fix AMD multi-tensor implementation.

jessebenson committed 5 years ago

245d4361

Disable Python 3.9 for training Python packaging build. (#6012)

edgchen1 committed 5 years ago

Verified 6572a4d3

longformer onnx conversion and benchmark tools (#6007)

tianleiwu committed 5 years ago

Verified cdb91208

Support Fusion for 1 and 2 Inputs Bert Models Converted From tf (#5993)

liuziyue committed 5 years ago

Verified 3b198c96

Fix Flaky orttraining tests (#5977)

SherlockNoMad committed 5 years ago

Verified c86a1e5c

Fix nuget build error (#6009)

Ryan Lai committed 5 years ago

Verified 2878e8eb

Fix build.py bug which prevents running some unit tests (#5990)

baijumeswani committed 5 years ago

Verified 2b35f7d4

Make operator TreeEnsemble 5x faster for batches of size 100.000 (#5965)

xadupre committed 5 years ago

Verified 0acc3837

Initialize a structure in operator ReduceSum (#6005)

xadupre committed 5 years ago

Verified 524b9fa8

Fix bugs for 1: Calibrator should check model inputs; 2: (#6017)

zhanghuanrong committed 5 years ago

Verified 648c9c77

Fix PR #5550 reverted in #5911 (performance improvment for operator Transpose) (#5916)

xadupre committed 5 years ago

Verified bdd06f63

Fuse MatMulIntegerToFloat only when scales are scalar (#6008)

yufenglee committed 5 years ago

Verified f2dcba7a

Add int32_t support for DeQuantizeLinear (#5994)

yufenglee committed 5 years ago

Verified 4fdfbfd4

include gemm_helper.h (#5988)

oliviajain committed 5 years ago

Verified c727a287

skip the check for A channel (#5989)

zhangxiang1993 committed 5 years ago

Verified b4e6cc59

[NNAPI] Support non-1d tensor for C of Gemm op (#5982)

guoyu-wang committed 5 years ago

Verified cdacee66

Use loose version in build.py (#5998)

guoyu-wang committed 5 years ago

Verified 6846c665

Add suspend handler with new telemetry event for UWP scenarios (#5907)

Ryan Lai committed 5 years ago

Verified 897310f6

Replace direct pulls from image cache container registry with get_docker_image.py, build definition clean up. (#5906)

edgchen1 committed 5 years ago

Verified 6d642a3d

Newer Older