2 ms·
Yes, i did that but its not apple silicon optimized so it was taking forever for 30b models. So its ok, but its not fantastic
by avetiszakharyan 1y ago
Yes, i did that but its not apple silicon optimized so it was taking forever for 30b models. So its ok, but its not fantastic
- spmurrayzzz 1y agoYou can just use llama.cpp instead (which is what ollama is using under the hood via bindings). Just need to make sure youre using commit `d3bd719` or newer. I normally use this with nvidia/cuda, but tested on my mbp and havent had any speed issues thus far. Alternatively, LMStudio has MLX support you can use as well.