The raw data used for training and validation can be found in this link (too large to host on GitHub). A trained version of the model can also be found in the same link for inference purposes.