iGerman00 / Pollyduble

An experimental proof-of-concept script to automatically dub videos to English with the help of local TTS, voice cloning, audio separation, speech recognition and FFmpeg
MIT License
8 stars 2 forks source link

Errors Installing & Usage #1

Open amjiddader opened 4 months ago

amjiddader commented 4 months ago

2 We can not install due to pip dependencies errors.

Still if we somehow manage to install it it does not work?

I am installing on Windows with Nvidia GPU and i tried several methods but i was not able to install due to pip deps issues. Somehow i managed to get pass using pip-tools now i have installed.

Its giving below error and the file "OpenVoice/checkpoints_v2/converter/config.json'" actually does not exist


Loading libraries... Initializing folders at C:\Users\Amjad\Videos\dub\Polydub\output Using device: cuda Traceback (most recent call last): File "C:\Users\Amjad\Videos\dub\demo.py", line 263, in main() File "C:\Users\Amjad\Videos\dub\demo.py", line 245, in main tone_color_converter = init_tone_color_converter() File "C:\Users\Amjad\Videos\dub\demo.py", line 62, in init_tone_color_converter tone_color_converter = ToneColorConverter(f'{config["CKPT_CONVERTER"]}/config.json', device=device) File "c:\users\amjad\videos\dub\src\myshell-openvoice\openvoice\api.py", line 103, in init super().init(*args, **kwargs) File "c:\users\amjad\videos\dub\src\myshell-openvoice\openvoice\api.py", line 21, in init hps = utils.get_hparams_from_file(config_path) File "c:\users\amjad\videos\dub\src\myshell-openvoice\openvoice\utils.py", line 7, in get_hparams_from_file with open(config_path, "r", encoding="utf-8") as f: FileNotFoundError: [Errno 2] No such file or directory: 'OpenVoice/checkpoints_v2/converter/config.json'

iGerman00 commented 4 months ago

That just shows that you haven't got the proper directory structure. You need to download the OpenVoice models - follow their instructions for the demos.

amjiddader commented 4 months ago

i got that, working now, thanks,

But it has simple TTS voice, i was expecting voice clone to clone the real looking voice from audio