search

keithito / tacotron

A TensorFlow implementation of Google's Tacotron speech synthesis with pre-trained model (unofficial)
MIT License
2.96k stars 957 forks source link

faster synthesis #246

Closed mmomin93 closed 5 years ago

mmomin93 commented 5 years ago

After training the model, the synthesis duration time of speech from text in local host server is very long. How to decrease the synthesis time on a PC system (RAM 8GB) ?

keithito commented 5 years ago

Run synthesis on a GPU, if available. Even a slow GPU will beat the CPU. You need to install a version of tensorflow that works with the GPU to do this. See https://www.tensorflow.org/install/gpu

MikeyBeez commented 4 years ago

Thank for supplying this info Keith. It's very good quality, but it's still really slow on a GPU. It's not usable for real-time TTS. I've been using gTTS, but it's getting to be rather broken these days. I really want a good local solution for TTS. espeak and festival stink, however. :(