Train-Custom-Speech-Model icon indicating copy to clipboard operation
Train-Custom-Speech-Model copied to clipboard

Published 20 hours ago • verified publisher icon IBM

Transcribe file > 15mb

Open dastevens05 opened this issue 5 years ago • 1 comments

The transcribe section won't accept a file >15mb. Do you have instructions for utilizing the trained models on an mp3 in a bucket or somewhere? My goal is to run this on a 1 hour audio file that is >15mb. Currently unable to do this in the GUI. Thanks.

dastevens05 avatar Sep 24 '20 13:09 dastevens05

something you could try:

  • increase the file size limitation here: https://github.com/IBM/Train-Custom-Speech-Model/blob/master/client/src/config.js#L4
  • convert your mp3 to wav and use python script here: https://github.com/IBM/Train-Custom-Speech-Model#6-transcribe-your-dictation

yhwang avatar Sep 24 '20 17:09 yhwang