alt Hacker NewsTraining is an overkill at this point imo. I have seen agents work quite well with a feedback loop, some tools and prompt optimisation. Are you doing fine-tuning on the models when you say training?
Training is an overkill at this point imo. I have seen agents work quite well with a feedback loop, some tools and prompt optimisation. Are you doing fine-tuning on the models when you say training?
Nope - just use memory layer with model routing system.
https://github.com/rush86999/atom/blob/main/docs/EPISODIC_ME...