I appreciated it, since he didn’t do a legit stress test. Running a local llm is intensive on the hardware, and if it performs well on that, it’ll likely perform well on most standard, non-useless tasks. So, I see that part as a makeshift stress test.