RTX 5090 local inference for Llama

[amberhall]

I'd approach it in a boring order. On 4090, the easiest win is often to back off the aggressive settings and see whether the problem survives. Not glamorous advice, but repeatable baselines save a lot of time.