saymrwulf/onnxruntime
1
0
Fork 0
mirror of https://github.com/saymrwulf/onnxruntime.git synced 2026-06-04 23:59:56 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions
onnxruntime/onnxruntime/python
History
stevenlix ce0025d3f2
Fallback Pow op in layer norm to FP32 in TRT to avoid overflow (#13639) 
Accuracy loss is observed when transformer models such as BERT, DeBERTa,
ViT are running in TRT FP16 mode. The cause is that overflow happens at
Pow op in layer norm.
This PR provides the option to force Pow to run in TRT FP32 precision if
overflow occurs.

Co-authored-by: Ubuntu <azureuser@orteplinuxdev.bxgbzpva45kedp3rhbsbit4phb.jx.internal.cloudapp.net>
2022-11-29 13:37:31 -08:00
..
backend replace 'master' branch ref to 'main' for onnx repo (#12678) 2022-08-30 13:41:42 -07:00
datasets replace invalid sample (#4567) 2020-07-21 23:51:17 -07:00
providers/tvm Format all python files under onnxruntime with black and isort (#11324) 2022-04-26 09:35:16 -07:00
tools Fallback Pow op in layer norm to FP32 in TRT to avoid overflow (#13639) 2022-11-29 13:37:31 -08:00
torch_cpp_extensions [ORTModule] ATen Support for aten::upsample_nearest (#13364) 2022-10-20 08:30:04 +08:00
training
__init__.py
_ld_preload.py Fix python manylinux to not load cuda if it fails to load dependencies (#8882) 2021-09-07 11:09:25 -07:00
_pybind_state.py.in Make ORT callable from various Pytorch compilers (LazyTensor, TorchDynamo, etc) (#10460) 2022-08-22 09:40:40 -07:00
exported_symbols.lst
numpy_helper.h Fix SDL warnings in CPU EP (#9975) 2021-12-19 20:54:29 -08:00
onnxruntime_collect_build_info.py Format all python files under onnxruntime with black and isort (#11324) 2022-04-26 09:35:16 -07:00
onnxruntime_inference_collection.py Enable ORT in TorchDynamo (#13259) 2022-11-01 11:19:29 -07:00
onnxruntime_pybind.h fix windows ci debug build break (#11495) 2022-05-12 16:54:00 -07:00
onnxruntime_pybind_exceptions.cc Python bindings fix ups in preparation to Sparse Tensor introduction (#7817) 2021-05-26 09:47:41 -07:00
onnxruntime_pybind_exceptions.h Fix issue with debug VS2022 build when python bindings are enabled (#9794) 2021-11-18 16:58:02 +10:00
onnxruntime_pybind_iobinding.cc Adds missing numpy type when looking for the ort correspondance (#10943) 2022-03-22 14:44:48 -07:00
onnxruntime_pybind_mlvalue.cc Switch GSL to MS GSL 4.0.0 (#13416) 2022-10-29 04:15:20 -07:00
onnxruntime_pybind_mlvalue.h Move OrtValueVector from onnxruntime-training to onnxruntime (#11176) 2022-06-15 09:36:28 +02:00
onnxruntime_pybind_module.cc Fix issue with debug VS2022 build when python bindings are enabled (#9794) 2021-11-18 16:58:02 +10:00
onnxruntime_pybind_ortvalue.cc Enable ORT in TorchDynamo (#13259) 2022-11-01 11:19:29 -07:00
onnxruntime_pybind_schema.cc Document generation for operator kernels, enable internal overload of DML EP to initialize on software-only devices (#13428) 2022-10-25 11:14:43 -07:00
onnxruntime_pybind_sparse_tensor.cc Adds missing numpy type when looking for the ort correspondance (#10943) 2022-03-22 14:44:48 -07:00
onnxruntime_pybind_state.cc Fallback Pow op in layer norm to FP32 in TRT to avoid overflow (#13639) 2022-11-29 13:37:31 -08:00
onnxruntime_pybind_state.h Fix issue with debug VS2022 build when python bindings are enabled (#9794) 2021-11-18 16:58:02 +10:00
onnxruntime_pybind_state_common.cc Allow CUDA EP enable or disable TunableOp via session options and environment variable (#13601) 2022-11-15 14:43:54 +08:00
onnxruntime_pybind_state_common.h Allow CUDA EP enable or disable TunableOp via session options and environment variable (#13601) 2022-11-15 14:43:54 +08:00
onnxruntime_validation.py Format all python files under onnxruntime with black and isort (#11324) 2022-04-26 09:35:16 -07:00
pybind.def
version_script.lds
version_script_expose_onnx_protobuf.lds Support external custom operator schemas on Ubuntu (#8807) 2021-08-28 11:05:21 -07:00