training pytorch models with onnxruntime

to build (you need to update version number in version.txt in order to be able to upload python whl):

rm dist/*
python setup.py bdist_wheel

to publish:

twine upload dist/*

to install:

stable:

(uninstall onnxruntime version that do not support gpu and training)
pip install torchort onnxruntime=1.9.0+cu111_training

nightly:

(make sure you are not in the ort repo folder - otherwise torchort is taken as already installed)
pip install --pre torchort ort-gpu-nightly-training

(eventually we are aiming at: pip install --pre torchort onnxruntime=1.9.0+cu111_training)

to test:

python ./ort/tests/bert_for_sequence_classification.py

to use torchort within PyTorch training scripts:

import onnxruntime
from torchort import ORTModule
model = ORTModule(model)
# normal PyTorch training script follows