How to use mio/tokiwa_midori with ESPnet:
from espnet2.bin.tts_inference import Text2Speech model = Text2Speech.from_pretrained("mio/tokiwa_midori") speech, *_ = model("text to generate speech from")
