NtTestAlert/tortoise-tts
1
0
Fork 0
forked from mrq/tortoise-tts
Code Issues Pull Requests Packages Projects Releases Wiki Activity

yeah

Browse Source
This commit is contained in:
James Betker 2022-04-18 10:30:22 -06:00
parent 3f968bedb5
commit a4bc51cb6d
1 changed files with 2 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
do_tts.py
View File

@ -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('--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!) ' 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') '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/') parser.add_argument('--output_path', type=str, help='Where to store outputs.', default='results/')
args = parser.parse_args() args = parser.parse_args()
os.makedirs(args.output_path, exist_ok=True) os.makedirs(args.output_path, exist_ok=True)
tts = TextToSpeech(autoregressive_batch_size=args.batch_size) tts = TextToSpeech()
voices = get_voices() voices = get_voices()
selected_voices = args.voice.split(',') selected_voices = args.voice.split(',')
@ -28,6 +25,6 @@ if __name__ == '__main__':
for cond_path in cond_paths: for cond_path in cond_paths:
c = load_audio(cond_path, 22050) c = load_audio(cond_path, 22050)
conds.append(c) 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) torchaudio.save(os.path.join(args.output_path, f'{voice}.wav'), gen.squeeze(0).cpu(), 24000)