Training data and preprocessing pipeline? #20
Description
At some point, a link was present in the README to download the training data, but it looks like it does not work anymore. Is the training dataset available somewhere?
In addition to the .dat file, it would be nice to have access to the raw training data and the preprocessing pipeline: the xml files, the midi files, and any other file and instructions needed to reproduce the training (e.g. the *_infer_corresp.txt files, how are they created, what is their format...)
I would like to train this model on some other data, so I would like to know how to process it in order to train the model with it.