search

antmicro-labs / voice-assistant

Apache License 2.0
0 stars 1 forks source link

Customizing model training parameters #25

Open zawadzki-dawid opened 4 years ago

zawadzki-dawid commented 4 years ago

Together with @plonajakub we are going to teach our model to recognize more words and we will focus on customizing the model parameters for lower latency while losing as little accuracy as possible. Here and here you can find more information about the model. I have also created a cheat sheet containing a list of parameters we can tweak during training.

rw1nkler commented 4 years ago

Hey, how is your progress in this task?

zawadzki-dawid commented 4 years ago

We retrained the model to recognize more words but we have not been playing with parameters yet. In the future we will try for sure to tweak some parameters and compare efficiency of the models. Now we have to focus more on presentation we have next week, we will try to design an app that can show capabilities of our model. @plonajakub should soon create new issue where you can find more info about the presentation. Retrained model can be found here.

szymon-filipiak commented 4 years ago

There was a problem with the mismatched version of TensorFlow in our project and version which was used for training.