Reduce processing time

[mrNicky]

If anyone, have any ideas on to reduce processing time. 20 secondes for me, between running command and saving wav output.
with a one line sentence.

I used default tts command like this:

tts --text "blabla wha. . .."

> tts_models/en/ljspeech/tacotron2-DDC is already downloaded.
 > vocoder_models/en/ljspeech/hifigan_v2 is already downloaded.
 > Using model: Tacotron2
 > Generator Model: hifigan_generator
Removing weight norm...
 > Text:  blablablablablablablablablablablablablablablablablablablablablablablablablablablablablabla
 > Text splitted to sentences.
 > Processing time: 8.45993447303772
 > Real-time factor: 0.6050886026393558
 > Saving output to tts_output.wav

[erogol]

how fast do you need? It is already faster than the real-time (length of the generated audio is longer than the execution time)
> Real-time factor: 0.6050886026393558

[L0CH1NBEK]

I also wondering, that is it possible to reduce processing time? the time between running command and getting the output .wav file.