Tacotron1 Deployment on t2.large

trying to deploy Mozilla tacotron1 model on 2 t2.large and one load balancer. I’m using Nginx but it is unable to handle parallel requests. The model is not handling multiprocessing.

server implementation is just for demoing. I don’t think it is efficient for deployment.

then how to deploy tacotron model?

you need to devise your own code for it, at least for now.