onnxruntime/onnxruntime
Hector Li a0ba25f98f
Fix the issue that it run into alloc failed on multiple cuda device. We have some place hard code the allocator always use device 0. (#1815)
Fix the issue that it run into alloc failed on multiple cuda device. We have some place hard code the allocator always use device 0.
2019-09-12 15:31:59 -07:00
..
automl_ops Implement C API entry points for creating and fetching non-standard types to OrtValue (#1714) 2019-09-11 14:52:47 -07:00
contrib_ops Fix the issue that it run into alloc failed on multiple cuda device. We have some place hard code the allocator always use device 0. (#1815) 2019-09-12 15:31:59 -07:00
core Fix the issue that it run into alloc failed on multiple cuda device. We have some place hard code the allocator always use device 0. (#1815) 2019-09-12 15:31:59 -07:00
python Enhance compatibility with proto3 and replace or abstract has_*() methods. (#1778) 2019-09-09 14:07:30 -07:00
server Bump onnx to latest (#1756) 2019-09-12 11:46:49 -07:00
test [Nuphar] improvements in symbolic_shape_infer and model editor (#1787) 2019-09-12 15:11:52 -07:00
__init__.py Python API naming and other cleanup (#1678) 2019-08-27 12:48:46 -07:00
ReformatSource.ps1
ReformatSourcePython.bat
VSCodeCoverage.runsettings