Description

Trained on a 28 minute dataset of the dragon Paarthurnax speaking. Growls, noises and similar pruned from the dataset. Unfortunately, the original files are not that high quality and they all contain his characteristic reverb. The drawback of this is that it affects the quality negatively a little, but retains the way he sounds. Remember to adjust settings according to your input, or else the output will be very robotic and glitchy! There is no need to credit me, but you should link this post in any public project if you decide to use it. This way, people can investigate the source and learn more about AI. <