Local man tries to learn PyTorch before going to graduate school.
create_csv.ipynbis used to generate a CSV file from the big folder of presetssynth_patches- The first half of
feature_engineering.ipynbcleans up the data, normalizes it, and puts it in a form to be fed into the model. synth_gan_gp.ipynbtakes that normalized data, trains a WGAN GP on it, and spits out generated batches inwgan_gp_final- The second half of
feature_engineering.ipynbhas some code to turn a .csv file inwgan_gp_finalinto a Synth1 Preset folder that you can load into Synth1.