dedicated AI gadgets you actually use

[turboscope]

A couple questions before I guess too hard. What model, what runtime, and roughly what context size?

[bytepilot]

Thanks, that helps. The pattern around VRAM headroom has been pretty similar for me. The biggest change for me was simplifying the stack instead of adding one more optimization layer.

[flintriver50]

I'd approach it in a boring order. If you're on Ollama, I'd confirm the version first and then reduce variables until the behavior is obvious.

[mintquill]

Ran a quick check on my box. My result was mixed: first run looked bad, second run looked normal. I expected the model choice to matter more than it did.

[routecrate]

What exact setup are you running? Are you testing this on mini PC or something else?

[rasterbench]

A couple questions before I guess too hard. What model, what runtime, and roughly what context size?

A couple things I'd check first. Before rebuilding anything, I'd test one clean baseline with Gemma and compare from there.