Jan Vainer
Jan Vainer
@adnan-mehremic Thanks for the info, I will try to replicate this during the weekend
Thanks for the link. My problem with this issue is that I am not able to reproduce this even with a clean setup and reinstalled dependencies and everything works even...
Thanks for trying this out! I will check if version 1.5.1 works for me and will bump up the requirement.
Hm it seems like you are right. This may be a bug. Did you compare the results with and without it?
This could maybe have some impact on the audio quality on the edges of the utterance. I noticed there some edge related problems before. Thanks for pointing this out!
Hi, thanks for trying this model out! Could you share the code adjustments you had to make in order to to train on your dataset? Could you please try to...
You can also resample your dataset with `sox` or `ffmpeg` and use the resampled audio too. @dsplog did your network train reasonably fast?
Good news :) Let me know how the training goes or if there are any issues.
Thanks for the update! The spectrograms will always be a bit blurry. I would not worry about it. Regarding inference, did you use the [inference script](https://github.com/janvainer/speedyspeech/blob/master/code/inference.py)? The provided MelGAN requires...
> I tested and found that the error occurs when running in the ctcdecode directory, but it will work in other directories. this worked for me too