Symbio: Self fine tuning ai agents

[huyedits]

I built Symbio, a local AI agent harness that automates the cycle of error --> solution loop on Apple silicon using MLX as well as MLX fine tuning LoRA. It uses a mixture of agents infrastructure where it triggers a LoRA update after five or the user threshold number amount of mistakes to dynamically keep on track on what you want it to do.
Thank you for reading!!
https://github.com/huyedits/Symbio