How long does it cost to train iKala data set on a single GPU?

andabi / music-source-separation

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

795 stars 150 forks source link

How long does it cost to train iKala data set on a single GPU? #21

Open youngsuenXMLY opened 6 years ago

youngsuenXMLY commented 6 years ago

Hi Andabi, I once used Matlab version of your project to separation speech from music, the result seemed promising. But Matlab version code cost too much time training. So I want to get some knowledge of how long this Tensorflow version costs to train, like iKala data set?

xrando commented 6 years ago

It would take about a few hours to a day or so depending on the length and duration of your dataset. Can i have the source codes of the matlab version you are using? thanks in advance!

youngsuenXMLY commented 6 years ago

Sorry for late reply. I used Posen Huang's matlab code. [https://github.com/posenhuang/deeplearningsourceseparation]

ucasiggcas commented 5 years ago

Have you got the ikala data ? Could you please share it ? Email: ucas010@126.com

Tortoise17 commented 5 years ago

Hi. Not yet. but if I will get. i will definitely share. I am also looking like you.

ucasiggcas commented 5 years ago

WHo can help me ?

No one !

It's a pitty.