chatterbox cli

import click
import torchaudio as ta
from chatterbox.tts import ChatterboxTTS
import sys

@click.command()
@click.option('--text', default=None, help='Text to synthesize (omit for stdin)')
@click.option('--exaggeration', default=0.5, help='Model exaggeration param')
@click.option('--cfg_weight', default=0.5, help='Model cfg weight param')
@click.option('--out', required=True, help='Path to the output file')
@click.option('--audio-prompt-path', default=None, help='Path to the audio prompt file')
def synthesize(text, exaggeration, cfg_weight, out, audio_prompt_path):
    model = ChatterboxTTS.from_pretrained(device="cuda")
    if text is None:
        text = sys.stdin.read()
    wav = model.generate(text, exaggeration=exaggeration, cfg_weight=cfg_weight, audio_prompt_path=audio_prompt_path)
    ta.save(out, wav, model.sr)
    print(f"Wrote {out}")

if __name__ == '__main__':
    synthesize()