You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi, I have finished training my own wake word model and exported .pb and .pb.params files. It worked very well, nice work guys, thank you!
But I want it to run on Android devices too. As far I as know, it should be in tflite format to be able to run on low energy chips on Android devices; so, questions is how can I do that? If it is not possible to convert to tflite, is there any trick to run the .pb model on Android OS?
The text was updated successfully, but these errors were encountered:
Try using the code from the tensorflow 2 PR. It includes support for tflite. Note, however, that the computation graph of the model expects MFCC features as inputs meaning after you get the tflite model running on Android you will need to port this computation.
Hi, I have finished training my own wake word model and exported
.pb
and.pb.params
files. It worked very well, nice work guys, thank you!But I want it to run on Android devices too. As far I as know, it should be in tflite format to be able to run on low energy chips on Android devices; so, questions is how can I do that? If it is not possible to convert to tflite, is there any trick to run the .pb model on Android OS?
The text was updated successfully, but these errors were encountered: