search

hkveeranki / speech-emotion-recognition

Speaker independent emotion recognition
http://speechemotionrecognition.repos.hkveeranki.com/
MIT License
315 stars 100 forks source link

I am facing a similar error in github. However I tried it in Google Colab and I got the following error: #12

Closed hkveeranki closed 5 years ago

hkveeranki commented 5 years ago

I am facing a similar error in github. However I tried it in Google Colab and I got the following error: NN has no attribute called 'train'. Could you please look into the issue? This is the Colab notebook. https://colab.research.google.com/drive/19zXD0GzR5-gXqoTEfZIW0PbHswdbc4ZD

Originally posted by @Nandy-Saran in https://github.com/harry-7/speech-emotion-recognition/issues/9#issuecomment-476726957

hkveeranki commented 5 years ago

@Nandy-Saran So, there are couple of things here.

  1. Model, MLModel and DLModel are abstract classes. Ideally you shouldn't be using them but instead use one of its child classes which implements all the methods and has the constructor.
  2. I made the package installable by providing a setup.py I don't know how it works in google colab but on a pc/laptop you should be install by running python setup.py install --local in your terminal from the package directory. This will enable to use this package wherever you want.
  3. Can you try using one of the given models or implement a new Model as child class to one of the above mentioned classes and then see if the training works?

Feel free to comment here if you have any other questions. You can also send me an email at harry7.opensource@gmail.com

hkveeranki commented 5 years ago

Closing this. Re open if you need anything