Ettore Di Giacinto

> whats best option here if we want to contribute just make forks of the branch and PRS against this? Yes, that would work just fine!

What is done: - [x] API spec - [x] Updating session, starting VAD server - [x] Hooking server API specs to placeholder functions - [x] Register ws server, and test...

Currently at creating the VAD backend with silero, attach it to the compilation process and to the binary releases

mh. things are in the good direction but still VAD isn't right, it detects the start of the conversation, but can't detect the end segment yet.

Extracted silero-vad bits over here: https://github.com/mudler/LocalAI/pull/4204 so can be tackled separately

closing as we merged it in https://github.com/mudler/LocalAI/pull/5392

> This is a good idea mudler... but it seems a _bit_ overactive. Can we tune that at all before we set it loose? Yes definitely, I've been opening this...

this issue is *not* refering to: https://github.com/kairos-io/kairos-sdk, but rather the the internal SDK for providers. The k3s provider consumes the new SDK logic, while the kairos provider doesn't: https://github.com/kairos-io/provider-k3s

WIP as need to still add mapping between the transformer tokenizer and the templates (see TODO note in code comments)

basic support should work (tested with llama3 prompt), probably is not going to cover all cases as gonja has limitations, but, since this kicks-in when no other template was defined...