You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I guess the pre-trained model that this project give is only trained for validation with only 4 data(p225, p226, p227, p228)
So to reproduce the result of paper, i download VCTK dataset and as paper say i pick 40 speaker (40 speakers for zero-shot train), which is p225 ~ p269 ( maybe 40 ~ 43 speaker, also i change sampling rate to 16k)
When making train.pkl with make_matadata.py there is error with len_crop.
ValueError: 'a' cannot be empty unless no samples are taken
The default value of len_crop is 128, but there is shorter data in VCTK.
I want to know whether you just remove the len_crop part or just use VCTK data which is longer then 128.
Thanks for reading issue.
The text was updated successfully, but these errors were encountered:
Thanks for your nice project and let us to use.
I guess the pre-trained model that this project give is only trained for validation with only 4 data(p225, p226, p227, p228)
So to reproduce the result of paper, i download VCTK dataset and as paper say i pick 40 speaker (40 speakers for zero-shot train), which is p225 ~ p269 ( maybe 40 ~ 43 speaker, also i change sampling rate to 16k)
When making train.pkl with make_matadata.py there is error with len_crop.
ValueError: 'a' cannot be empty unless no samples are taken
The default value of len_crop is 128, but there is shorter data in VCTK.
I want to know whether you just remove the len_crop part or just use VCTK data which is longer then 128.
Thanks for reading issue.
The text was updated successfully, but these errors were encountered: