End-to-End Text-to-Speech Synthesis using Transformer-Based Acoustic Modeling and Parallel Wave GAN