I had a version that on validation dataset nearly 1:1 across entire spectrum, but I ended up linking this revision more.
Also had qwen3-tts version, but no matter what I do it sounded more like brat Cortana. The way piper-tts work under the hood make it sound more machine-like.
Not a fan of java/spring boot, although kotlin is nice.
Currently all of my greenfield work projects are using alpine and tauri (rust). It's more complex than the django web apps, but already seeing strong performance gains and better runtime guarantees.
I've built a couple non-trivial, cross-platform apps with alpine.js and htmx with tauri as the sole backend with ipc. Claude is very good at full-stack with the tauri mcp driving harder e2e tests that leverage platform APIs.
I used gram predominately on the macbook neo and it worked well as a low-spec devbox
reply