WaveRNN
WaveRNN copied to clipboard
Unconditional weights / training tips?
Has anyone had success training the unconditional generation model? Any checkpoints or idea of how many epochs it takes to converge?
I trained an nc
model to ~150k steps and definitely heard what sounded like the faint murmur of my dataset… but nothing beyond that, really. The log reported an infinitesimal loss, but I don't think I believe it. In any case, just wanted to let you know there are others out here trying to get this working!
thanks! will probably not be trying nc waveRNN then, due to limited budget. But I'll post here if I do.