microsoft/onnxruntime

Pull Requests Commits

Fix CUDA provider factory include file

Pranav Sharma committed 4 years ago

718e5ab5

Fix bad URL to huggingface onnx-export example notebook

Dmitry Yutkin committed 4 years ago

067759b3

Enable cast propagation with level one by default. (#8286)

satyajandhyala committed 4 years ago

Verified 84bc20fe

Replace functions with secured version for OSX compliance (#7586)

RandySheriffH committed 4 years ago

Verified f40df302

autograd function fallback perf (#8312)

pengwa committed 4 years ago

Verified 6dbfb8db

Fix issue with ONNX to ORT format model conversion script when given single model file as input. (#8323)

edgchen1 committed 4 years ago

Verified c254c3c3

Support lists as inputs to ORTModule (#8311)

baijumeswani committed 4 years ago

Verified 6652d17d

Make Torch CPP extension build optional for packaging pipelines (#8305)

Thiago Crepaldi committed 4 years ago

Verified 9a855fe9

handle unsqueeze change in opset13 (#8308)

souptc committed 4 years ago

Verified d7c37033

Autograd Function Fallback bug fix - moe support (#8105)

pengwa committed 4 years ago

Verified 2347a0ac

Drop unused functions from math.h (#8304)

Nick Kreeger committed 4 years ago

Verified 40e5279f

Move kernel implementations outside of lookup table utility functions. (#8306)

Nick Kreeger committed 4 years ago

Verified 62d1458e

Pinning pillow version to 8.2.0 to circumvent regression introduced by 8.3.0 (#8303)

baijumeswani committed 4 years ago

Verified 090bae21

Use single builder image across Python versions for ROCm wheels (#8302)

Suffian Khan committed 4 years ago

Verified 008c5f76

Fix optimizer crash (#8274)

RandySheriffH committed 4 years ago

Verified 56e4dd1d

fix ld_preload for rocm (#8290)

Suffian Khan committed 4 years ago

Verified e71846b0

register softmaxinternal with rocm (#8289)

Suffian Khan committed 4 years ago

Verified 036eee5b

Update issue template to ask users to check known issues to avoid repetition. (#8288)

Pranav Sharma committed 4 years ago

Verified 969eb545

Remove path from telemetry strings (#8281)

tiagoshibata committed 4 years ago

Verified 0fa9ac36

Fix lamda function formatting in layer_norm.cc (#8276)

Nick Kreeger committed 4 years ago

Verified 552806f3

Pin version of Pillow to 8.2.0 to circumvent noncompatibility with numpy (#8278)

baijumeswani committed 4 years ago

Verified 2bda2a62

Support OrtMemTypeCPUInput for ATenOp/ATenOpGrad (#8116)

Lafi7e committed 4 years ago

Verified 88ec95ea

Add iOS packaging pipeline (#8264)

edgchen1 committed 4 years ago

Verified b42e7d2c

disable computation reduction optimization for non-gpu build (#8251)

souptc committed 4 years ago

Verified a9a2394f

enable BN training in cpu inference build (#8269)

Lafi7e committed 4 years ago

Verified 9cfe642b

fix the shared provider test for training build; expose more symbols to non cuda build (#8249)

souptc committed 4 years ago

Verified 996a98b3

Integrate onnxruntime-extensions into onnxruntime. (#8143)

Zuwei-Zhao committed 4 years ago

Verified b46310b3

Provide torch module interface for ORTModule (#8148)

baijumeswani committed 4 years ago

Verified f616cd07

gather elements optimization (#8154)

Lafi7e committed 4 years ago

Verified ce9d1349

ScatterNDGrad (#8261)

Lafi7e committed 4 years ago

Verified ef8f50c4

Older