Hi everyone
here some very informative article on training large language models (LLMs) which can be interesting for some fellow learners since it also touches upon a modern tech stack. Feel free to check out how Replit built their own LLM from scratch:
Best regards
Christian