search

fgnt / sms_wsj

SMS-WSJ: Spatialized Multi-Speaker Wall Street Journal database for multi-channel source separation and recognition
MIT License
101 stars 23 forks source link

Open Source Trained model #6

Closed hangtingchen closed 4 years ago

hangtingchen commented 4 years ago

Hi, Is there any possibility to make the already trained model open source? Since my research is mainly on the front-end but I have found that the training costs a lot of time and our machine is not enough.

jensheit commented 4 years ago

Thanks for your interest in SMS-WSJ. We cannot publish a model trained on WSJ data for licencing reasons. This holds for our acoustic model as well. However, if the model does not fit you GPU, you can look at the NN training here and reduce the batch size in line 149 or change the model definition (line 104-113).

If there are other questions regarding the am training please open a new issue.

hangtingchen commented 4 years ago

Thanks, I am trying,