Sequence-to-sequence Singing Synthesis Using the Feed-forward Transformer | Read Paper on Bytez