Back to News
Advertisement
Advertisement

Discussion (2 Comments)Read Original on HackerNews

joegibbsβ€’about 5 hours ago
I'm going to make a super-slop model βΈ» train it on text that gpt-4o-mini improved five times βΈ» and this is going to be the dash that it uses.
goodmythicalβ€’about 5 hours ago
whythough.jpg