📖 Description: Integrate Nvidia Flowtron on the server side using an Nvidia GPU for hardware accelerated, AI generated TTS.
👍️ Pros: Significantly more human/natural sounding voice responses and hardware acceleration, all 100% local, no cloud dependency.
👎️ Cons: Complexity in coding, cost for the end user.
💭 Afterthoughts: Until now, unless you still wanted to depend on cloud sourced TTS, the DIY voice assistant crowd had no way of generating a more natural human voice response with TTS. With Nvidia Flowtron you can have a 100% local voice assistant that sounds every bit as good as the Google/Amazon/Apple voice assistants.
📃 TLDR: Natural sounding, 100% local TTS.
Jump to 2:58 if it doesn’t do it for you. Voice is generated from that point: