Open japita-se opened 4 years ago
Yes, sure. But you will need to tune hyperparameters for each new architecture.
It would be great if someone would join the adventure :) my goal is to run deepspeech on a constrained board. I would try XNOR-net, but it seems that lcnn is superior.
Theoretically, It would feasible to use the same concepts for more complex architecyures like tacotron or deepspeech?