It’s been ok. I’m running llama2 7b and it’s … fine. The results I get from gpt4 aren’t much better. This for general tasks.

Mostly I think I need to use LLMs more effectively