2 articles · Updated 2026-10-01

Pit stop

Quick fixes for the problems everyone hits: out of memory, slow generation, picking a quant.

Which quant should I pick?

2026-10-01

Q4_K_M, IQ4_XS, Q6_K: what the names mean and where quality starts to drop.

Out of memory: what to try, in order

2026-10-01

The model does not load, or crashes on a long prompt. Seven fixes, from the cheapest to the most costly.