A simple example of training a Tensorflow model with Python, check the model with tract python API, then loading it into tract and compare predictions.
Setup environment.
pip install -r requirements.txt
Train a model, export it to ONNX along with a input and output example.
python example.py
(Outputs are commited to git, you don't need to run the python step at all.)
Run the model and double check the output.
cargo run