Anyone successfully running LLMs locally and using it on a daily basis? I just discovered the AI 395 Max exists a few days ago. https://x.com/AmadeusSVX/status/1953294265922048026 Someone managed to run GPT-OSS-120B at 30 tokens/sec on a tablet form factor device that has 128GB of RAM. This specific device seems to be going out of stock pretty fast.
I am just waiting for MINISFORUM's AI 395 pc, MS-S1 MAX, which will be released very soon.
I mean, you can go smaller than that... https://www.youtube.com/watch?v=VaeI9YgE1o8 (building an LLM in Minecraft)