They train LLM only with data up to 1930 and it still solves Python problems (theregister.com) 4 points by Multipassionate 2mo ago ↗ HN
1 comment
[ 4.4 ms ] story [ 10.8 ms ] thread