search

CorentinJ / Real-Time-Voice-Cloning

Clone a voice in 5 seconds to generate arbitrary speech in real-time
Other
52.6k stars 8.8k forks source link

[Errno 2] No such file or directory: 'encoder/_sources.txt' #250

Closed rkdlight closed 4 years ago

rkdlight commented 4 years ago

I have this problem, but I can't understand what does this file contain? There is not _sources.txt in this repo

rkdlight commented 4 years ago

Sorry, I dont describe my problem. I try to train this on Russian. But when I run encoder_train.py I get this error

zhudu commented 4 years ago

@CorentinJ I have the same problem, when I run encoder_train.py, it shows that the _source.txt file cannot be found.I want to know if this file was written by myself, or should it be in the database?And in your code, the file is added in encoder / data_objects / speaker.py. The database I downloaded does not contain such a file. If we need to write it ourselves, we hope you can give us the way to write it, which is helpful for the subsequent training of encoder.

rkdlight commented 4 years ago

You can have this problem because: 1) you don't run encoder_preprocess.py 2) if you use data is not from the librispeech or voxceleb, your data can have different format. 3)last: if preprocessing is finished successfully, but you have this problem, you are writing command wrong. It must look like this: encoder_train.py <name of your encoder model(it can be anything)> <directory which was made by preprocessing(find it)>

zhudu commented 4 years ago

Thank you very much. I did run encode_preprocess.py, but still reported an error. Maybe there is a problem with my data. I will see the reason from the data. Is your _source.txt in voxceleb1?