Skip to content
Better HN
They train LLM only with data up to 1930 and it still solves Python problems | Better HN