25 ms·
Running Qwen3:30B MoE on an RTX 3070 laptop with Ollama
- jdboyd 1y agoI think there is a lot of value in people documenting what it takes to make models that will run on 8, 12, and 16GB GPUs work well for them, and the tools used.
loading thread…