HarkonCollider/tortoise-tts
1
0
Fork 0
forked from mrq/tortoise-tts
Code Pull Requests Activity

Support combining voices in do_tts

Browse Source
This commit is contained in:
James Betker 2022-05-22 05:28:15 -06:00
parent 412315ab7d
commit 42a3bc9cfd
1 changed files with 8 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
tortoise/do_tts.py
View File

@ -5,7 +5,7 @@ import torch
import torchaudio import torchaudio
from api import TextToSpeech, MODELS_DIR from api import TextToSpeech, MODELS_DIR
from utils.audio import load_voice from utils.audio import load_voices
if __name__ == '__main__': if __name__ == '__main__':
parser = argparse.ArgumentParser() parser = argparse.ArgumentParser()
@ -25,8 +25,13 @@ if __name__ == '__main__':
tts = TextToSpeech(models_dir=args.model_dir) tts = TextToSpeech(models_dir=args.model_dir)
selected_voices = args.voice.split(',') selected_voices = args.voice.split(',')
for k, voice in enumerate(selected_voices): for k, selected_voice in enumerate(selected_voices):
voice_samples, conditioning_latents = load_voice(voice) if '&' in selected_voice:
voice_sel = selected_voice.split('&')
else:
voice_sel = [selected_voice]
voice_samples, conditioning_latents = load_voices(voice_sel)
gen, dbg_state = tts.tts_with_preset(args.text, k=args.candidates, voice_samples=voice_samples, conditioning_latents=conditioning_latents, gen, dbg_state = tts.tts_with_preset(args.text, k=args.candidates, voice_samples=voice_samples, conditioning_latents=conditioning_latents,
preset=args.preset, use_deterministic_seed=args.seed, return_deterministic_state=True) preset=args.preset, use_deterministic_seed=args.seed, return_deterministic_state=True)
if isinstance(gen, list): if isinstance(gen, list):