Training large language models is a very technically demanding and resource intensive task. Many such models also require lots of specialized compute to use. Fortunately, we have recently been blessed by public releases of several "small" LLMs 1/4

→ View original post on X — @tunguz,