Indeed. For bespoke behaviors, a lot of open models allow LoRA training or fine-tuning. It's really feasible for particular use cases, but it will require more processing power and some training.

Was this answer helpful? 0 Users Found This Useful (0 Votes)