Complete preprocess file plus add the data preprocessed.
Note: I decided to go with .pt format for mfcc and covariance. These were unfortunately too big to push to git, so their previous .parquet df implementations were added instead. The .pt are easily generated again by running preprocess with the GTZAN dataset downloaded locally
Complete preprocess file plus add the data preprocessed.
Note: I decided to go with
.pt
format for mfcc and covariance. These were unfortunately too big to push to git, so their previous.parquet
df implementations were added instead. The.pt
are easily generated again by running preprocess with the GTZAN dataset downloaded locally