Comment by ricardobayes
5 months ago
You can run it, however those low quantized models (iQ2, iQ4, Q2) will very likely underperform the 9B versions at Q6/Q8.
5 months ago
You can run it, however those low quantized models (iQ2, iQ4, Q2) will very likely underperform the 9B versions at Q6/Q8.
Something about qwen models hold up really well even at low quants. for most other models anything under q5 is cooked, but on 35B-A3B I can get a lot of things done even at q3_xl. It is definitely better than full precision 9B