Rendered at 07:51:34 GMT+0000 (Coordinated Universal Time) with Cloudflare Workers.
p_stuart82 10 hours ago [-]
Does this catch intermittent fallback between requests?
logickkk1 9 hours ago [-]
Yes, picchio monitors a running llama-server on a timer and flags when the prefill/decode ratio goes CPU-shaped. I hot-swapped one mid-run. probe 4 caught ENGAGED -> NOT ENGAGED.