I bought a Mac Mini specifically for experimenting with LLMs. On the rare times I want to try fine tuning, I use Colab Pro or rent a server on Lambda Labs. I ju

I bought a Mac Mini specifically for experimenting with LLMs. On the rare times I want to try fine tuning, I use Colab Pro or rent a server on Lambda Labs. I just use the M2 for inferencing, so far. So, no heat problems. EDIT: maybe the YouTube demo was using 2 bit quantization?