search

TzuchengChang / NASS

Noise-Aware Speech Separation with Contrastive Learning
14 stars 6 forks source link

How to Use My Own Data with the Pretrained Model #6

Open purplestrawberrywatermelon opened 1 month ago

purplestrawberrywatermelon commented 1 month ago

Hi, Thank you for providing this pretrained model. I would like to use this model with my own dataset, but I'm not sure how to properly input my data and obtain results. Could you please provide a step-by-step guide or point me to the relevant documentation?

TzuchengChang commented 1 month ago

To use your own data, pls refer to the speechbrain/recipes/LibriMix/separation/hparams/sepformer-libri2mix-custom.yaml. Your data should be organized as shown in the custom folder in readme. Specifically, if you just want to infer, then the noise folder can put the same number of files (just placeholder), if you want to train, then noise data is indispensable.

purplestrawberrywatermelon commented 1 month ago

Thank you for your detailed explanation! I appreciate your help.