"""

Dumps the conditioning latents for the specified voice to disk. These are expressive latents which can be used for

other ML models, or can be augmented manually and fed back into Tortoise to affect vocal qualities.

"""

from pathlib import Path

from typing import Literal

import torch

from simple_parsing.decorators import main

from tortoise.api import TextToSpeech

from tortoise.utils.audio import get_voices, load_required_audio

EXAMPLE = Path(__file__ + "/../../results/conditioning_latents").resolve()

@main

def main(

voice: str = "pat2",

output_path: Path = EXAMPLE,

latent_averaging_mode: Literal[0, 1, 2] = 0,

):

"""Dumps the conditioning latents for the specified voice to disk. These are expressive latents which can be used for

other ML models, or can be augmented manually and fed back into Tortoise to affect vocal qualities.

Args:

voice: Selects the voice to convert to conditioning latents

output_path: Where to store outputs.

latent_averaging_mode: How to average voice latents, 0 for standard, 1 for per-sample, 2 for per-minichunk

"""

output_path.mkdir(exist_ok=True)

tts = TextToSpeech()

voices = get_voices()

selected_voices = voice.split(",")

for voice in selected_voices:

cond_paths = voices[voice]

conds = []

for cond_path in cond_paths:

c = load_required_audio(cond_path)

conds.append(c)

conditioning_latents = tts.get_conditioning_latents(

conds, latent_averaging_mode=latent_averaging_mode

)

torch.save(conditioning_latents, output_path / f"{voice}.pth")

if __name__ == "__main__":

main()