VCTK-VITS Finetune #929
Unanswered
StephanosPSteer
asked this question in
General Q&A
Replies: 1 comment
-
Just to say I worked out the is a directory error but now just as training starts I am getting this error TypeError: conv1d() received an invalid combination of arguments - got (NoneType, Parameter, Parameter, tuple, tuple, tuple, int), |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Like the discussion at
#854
I am having problems with finetuning VCTK-VITS
I am following the recipe https://github.com/coqui-ai/TTS/blob/main/recipes/vctk/vits/train_vits.py and have changed the outputpath but I was still getting errors. I traced them back and the Formatter function VCTK in formatters.py which seemed to expect multiple meta files.
I created my own formatter function took loads from the ljspeech one and it started to split the file
items = []
but I got the is a directory error
IsADirectoryError: [Errno 21] Is a directory: '/home/ubuntu/.local/share/tts/tts_models--en--vctk--vits'
I am pretty new to python (but not programming) and very new to TTS so I don't really want to go too far down the rabbit hole. I attached my meta file in case that will help but can you advise on where I go from here. Is the VCTK formatter ok? would my one work if there is only one extra speaker?
Thank you
s
Meta.txt
Beta Was this translation helpful? Give feedback.
All reactions