File size: 1,473 Bytes
f71bc95 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 | from __future__ import annotations
import argparse
from .infer import load_generator, write_wave
def build_parser() -> argparse.ArgumentParser:
parser = argparse.ArgumentParser(prog="wfloat-tts")
parser.add_argument("--model", "--checkpoint", dest="model", default="model.safetensors")
parser.add_argument("--config", default="config.json")
parser.add_argument("--text", required=True)
parser.add_argument("--sid", type=int, default=0)
parser.add_argument("--emotion", default="neutral")
parser.add_argument("--intensity", type=float, default=0.5)
parser.add_argument("--noise-scale", type=float, default=None)
parser.add_argument("--length-scale", type=float, default=None)
parser.add_argument("--noise-w", type=float, default=None)
parser.add_argument("--device", default="cpu")
parser.add_argument("--output", required=True)
return parser
def main() -> None:
parser = build_parser()
args = parser.parse_args()
generator = load_generator(
checkpoint_path=args.model,
config_path=args.config,
device=args.device,
)
audio = generator.generate(
text=args.text,
sid=args.sid,
emotion=args.emotion,
intensity=args.intensity,
noise_scale=args.noise_scale,
length_scale=args.length_scale,
noise_w=args.noise_w,
)
write_wave(args.output, audio.samples, audio.sample_rate)
if __name__ == "__main__":
main()
|