I got it working; just had to adjust the dataset in the JSON file to match the filename: changed from “en_US-fedcomp_2429” to “en_US-fedcomp-medium”.
I also adjusted the length_scale to 1.1 to try and slow the voice down a little.
1 Like