This is a standard Tacotron2 Text-to-Speech model based on the character Brawn from Transformers Generation 1.  This model was trained on 72 sample wavs for a total of 6 minutes and 29 seconds of audio data.