From a4bc51cb6db0d3f662a8bea56d2769d32f184f91 Mon Sep 17 00:00:00 2001 From: James Betker Date: Mon, 18 Apr 2022 10:30:22 -0600 Subject: [PATCH] yeah --- do_tts.py | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-) diff --git a/do_tts.py b/do_tts.py index ec21641..611d698 100644 --- a/do_tts.py +++ b/do_tts.py @@ -11,14 +11,11 @@ if __name__ == '__main__': parser.add_argument('--text', type=str, help='Text to speak.', default="I am a language model that has learned to speak.") parser.add_argument('--voice', type=str, help='Selects the voice to use for generation. See options in voices/ directory (and add your own!) ' 'Use the & character to join two voices together. Use a comma to perform inference on multiple voices.', default='patrick_stewart') - parser.add_argument('--num_samples', type=int, help='How many total outputs the autoregressive transformer should produce.', default=256) - parser.add_argument('--batch_size', type=int, help='How many samples to process at once in the autoregressive model.', default=16) - parser.add_argument('--num_diffusion_samples', type=int, help='Number of outputs that progress to the diffusion stage.', default=16) parser.add_argument('--output_path', type=str, help='Where to store outputs.', default='results/') args = parser.parse_args() os.makedirs(args.output_path, exist_ok=True) - tts = TextToSpeech(autoregressive_batch_size=args.batch_size) + tts = TextToSpeech() voices = get_voices() selected_voices = args.voice.split(',') @@ -28,6 +25,6 @@ if __name__ == '__main__': for cond_path in cond_paths: c = load_audio(cond_path, 22050) conds.append(c) - gen = tts.tts(args.text, conds, num_autoregressive_samples=args.num_samples) + gen = tts.tts_with_preset(args.text, conds, preset='standard') torchaudio.save(os.path.join(args.output_path, f'{voice}.wav'), gen.squeeze(0).cpu(), 24000)