Model persistence

[spestana]

It seems that because we were using joblib files to save prior versions of the trained planetsca model, this means that the model can only be used in an environment that matches the one in which it was trained (same package version, I believe in our case it was a scikit-learn 1.3 version)

Here is information about model persistence for scikit-learn models, where you can see that joblib models "Requires the same environment as the training environment"

Of the options listed there, the ONNX is the only one that doesn't rely on a specific python environment. We might want to look into converting the joblib models into ONNX formats.

[spestana]

This has been partially addressed with #36 , but I'm going to leave this issue open for now. The ONNX functions I've added need more testing (as well as the ONNX version of the models on Hugging Face).