Comment by simonw

"Fine-tuning large language models (LLMs) is frequently sold as a quick, powerful method for injecting new knowledge"

Is that true though? I don't think I've seen a vendor selling that as a benefit of fine-tuning.