Niklss

You know, I just decreased the sample rate and it worked for me pretty well!!! `sf.write(filename, generated_wav.astype(np.float32), round(synthesizer.sample_rate / 1.1))` At least it reduced the noise =)

> @Hubotcoder, tried contacting the authors via mail but no reply even there. How did you advance with this repository? Were you able to train the models? > > Thank...

> _No description provided._ Is it still relevant? My checkpoint weights 8.5GB, don't know an easy way to share =)

> I also meet this issue. How to evaluate with test dataset? I simply rewrote the code to also evaluate on test and save the best model based on test...

> How much GPU memory is required for model training? If the GPU memory is not enough, how can model parameters be optimized (batch_size has been set to 1)? I've...