QLoRA resource reality check with LM Studio

[cacheline]

[quote]Thanks, that helps. The pattern around continuous batching has been pretty similar for me. I got better results once I stopped chasing one metr

I'd approach it in a boring order. Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at prefill/decode split. Not glamorous advice, but repeatable baselines save a lot of time.

[deltastation]

Appreciate the detail. The pattern around batch size has been pretty similar for me.

[quiethouse]

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.