I see a lot of tutorials out there about how to fine-tune LLMs, but not a lot on how to prepare the data. These researchers at Snorkel AI explain how they curated the data, and what the impact was.
I see a lot of tutorials out there about how to fine-tune LLMs, but not a lot on how to prepare the data. These researchers at Snorkel AI explain how they curated the data, and what the impact was.