SerafimC

To stop usually I just kill . To restore you have to use the argument --restore_from and appoint the last version of yout model at the logdir folder.

It´s sad that the author does not maintains this repository anymore. Is it possible to add new reviewers that could aprove these changes?

Just to share my experience, I am using a dataset with one sentence per voice sample. So the audios have around 3 to 5 seconds each.

You can try with any voice dataset. Currently Im testing with my one recorded voice.

@PascalSchimmler thank you for the information. I was looking for some notion of how many steps is necessary to create good outputs.

Try to decrease trim_silence threshold with the parameter --silence_threshold Im testing currently with --silence_threshold=0.1

This still needs a fix. Any update?