Training and integrating GPT and other language models to build a voicebot system
D. -T. Nguyen, N. -Q. Tran, V. -K. Pham, S. -T. Wu · IET conference proceedings. · 2024
The paper proposes the integration of multiple models to build a voicebot. We train the GPT model to generate responses from text input from the user, and other models to process the user's audio signal into text and to transform the text output from the GPT model into an audio signal.