Thiago Crepaldi f3b0c93a45 Fix issue preventing loss scaler to run due (#4833) ... `LossScaler.update()` was not being properly called due to the incorrect TrainStepInfo.all_finite assignment. Additionally to this fix, _ORTTrainerModelDesc.is_finite was renamed to _ORTTrainerModelDesc.all_finite to make it more uniform with TrainStepInfo		2020-08-18 10:03:02 -07:00
..
orttraining	Fix issue preventing loss scaler to run due (#4833)	2020-08-18 10:03:02 -07:00
pytorch_frontend_examples	Remove ORTModel from frontend API (#3825)	2020-05-11 18:20:33 -07:00
tools	Update convergence baseline for ci_test. (#4465)	2020-07-09 15:29:36 +08:00