A GLaDOS TTS, using Forward Tacotron and HiFiGAN. Inference is fast and stable, even on the CPU. A low quality vocoder model is included for mobile use. Rudimentary TTS script included. Works perfectly on Linux, partially on Maybe someone smarter than me can make a GUI.
Added docker file and docker ignore to allow usage with docker. public image can be found at illusivebagel/glados-tts