search

alphacep / vosk-android-demo

Offline speech recognition for Android with Vosk library.
Apache License 2.0
714 stars 187 forks source link

In the model after training there isn't ivector folder. #209

Open chungtran2000 opened 10 months ago

chungtran2000 commented 10 months ago

After successfully running the file train.py I get the files mfcc.conf, final.mdl,HCLG.fst, words.txt, .... But compared to the folder downloaded from "https://alphacephei.com/vosk /models" I see the ivector folder is missing. And when I put it into the Android project I got an error: Failed to create model. What do I need to do to fix this error? github

nshmyrev commented 10 months ago

Which train.py do you run?

chungtran2000 commented 10 months ago

I made the model following this tutorial: "https://github.com/undertheseanlp/automatic_speech_recognition". The train.py file that I run is located at the project's path:"automatic_speech_recognition/egs/vivos".

nshmyrev commented 10 months ago

That setup is not sufficient. Our training demo is here, you need to follow it instead:

https://github.com/alphacep/vosk-api/tree/master/training

chungtran2000 commented 10 months ago

Thank you so much. Do you have specific instructions for this?

nshmyrev commented 10 months ago

No instructions, you can just run it locally or in colab. Let us know if you have any questions