LoRA works well for diffusion image generation, but I haven't had much luck with LLMs. I also don't have the resources to train a larger LLM
yeah building a good training data set seems to be the challenge
Please Login to reply.
Yes, in my experience, this is what takes the most time. There is a lot that you can automate, but there is still a lot that takes a human touch if you want good results.
