LLMs need a LOT of #data to train on. This is the Common Crawl dataset, which is one of the few that are free to the public. Check out how big just the July 2024 dataset is. Jeez. Link in reply.