3 ms·

I tried an hour ago and had a can't load model error. Everything up to date. Is there any special step?

• mmcclure 3y ago
  Tried `ollama pull mixtral` just now and it seems to be working, albeit pretty slowly.
• Casteil 3y ago
  How much RAM do you have? Mixtral is a beast and the non quantized model wants 40GB+ of memory.
  • eurekin 3y ago
    Ah, that might be it! I have only 32
    • ukuina 3y ago
      The q2 should fit.