The train and test csv files refer to hardcoded paths

[mazz]

cocktail-fork-test.csv
cocktail-fork-train.csv

both have paths to wav files that are hardcoded

ex:

D:\dnr\cv\10084\annots.csv,D:\dnr\cv\10084\mix.wav,D:\dnr\cv\10084\music.wav,D:\dnr\cv\10084\sfx.wav,D:\dnr\cv\10084\speech.wav

[ksasso1028]

Hello @mazz

I recommend running the make_dataset.py script and changing the path to your local train/test paths for audio

audio-reverb-removal/make_dataset.py

Line 11 in 9382b21

train = ["D:\\dnr\\tr\\"]