Comment by aurareturn

This article keeps getting posted but it runs a thinking model at 3-4 tokens/s. You might as well take a vacation if you ask it a question.

It’s a gimmick and not a real solution.