How to split the datasets for training/test/validation if I only have mir-1k ?

andabi / music-source-separation

Deep neural networks for separating singing voice from music written in TensorFlow

795 stars 150 forks source link

Open ucasiggcas opened 5 years ago

ucasiggcas commented 5 years ago

Dear,

If I only have the mir-1k,No ikala dataset, how could I modify the config.py ?

How to prepare the dataset ?

Any one who have run the project success please help me ?

SOS Thx

ucasiggcas commented 5 years ago

how to set the config.py ? Who can help me ? Please.

Who can reproduce the project ?

ucasiggcas commented 5 years ago

@andabi

ucasiggcas commented 5 years ago

Any one runs success the project ?