search

andabi / music-source-separation

Deep neural networks for separating singing voice from music written in TensorFlow
796 stars 149 forks source link

Ikala config #42

Open Tortoise17 opened 5 years ago

Tortoise17 commented 5 years ago

HI.

I would like to learn here, I need your help and guide that what changes to be done if use iKala or any other dataset. I am specially interested in iKala config.

Please your kind words will be appreciated.

ucasiggcas commented 5 years ago

ikala will not supply, and I find the batch_size=1, this is not good, do u have the train logs ? thx