power delivery for 24/7 serving

[vramport]

that fixed it for me too.

[batchwire]

What exact setup are you running? What model, what runtime, and roughly what context size? Turns out 'just one more test' was a lie.

[homelabhenry]

What exact setup are you running? What model, what runtime, and roughly what context size? Turns out 'just one more test' was a lie.

My first guess would be this: Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at quant drift. Not glamorous advice, but repeatable baselines save a lot of time.

[stackpilot]

I hit almost the same issue last night.

[homelabhenry]

What exact setup are you running? Is this local-only or are you routing anything externally?