ylying
ylying
just use MFA to generate textgrid files and resample to same sampling rate is OK，its much better than aishell
yes, as I tested, set max seq_len 80 is good
> > yes, as I tested, set max seq_len 80 is good > > @yileld if I have trained a model, do I need to re-train it to apply the...
> > yes, as I tested, set max seq_len 80 is good > > it seems worked,and why? just because training datasets has no long sentence lol
> > > > yes, as I tested, set max seq_len 80 is good > > > > > > > > > it seems worked,and why? > > >...
the best solution is to run "mfa train xxx" command to generate textgrid files again and then run preprocess.py(even run "mfa align xxx" command using align model trained in other...
when you train a model with multi GPU, state_dict will be saved as model.module.state_dict while model.state_dict when single GPU.
run “mfa train” instead of “mfa align”
the best solution is to run "mfa train xxx" command to generate textgrid files again and then run preprocess.py(even run "mfa align xxx" command using align model trained in other...
> I just found out I have some file listed in `unaligned.txt` in TextGrid folder, should I delete them? it doesnt matter, just keep it. lexicon and lab no need...