I want to
how do I run an LLM on my own computer and use it offline
Offline means the weights and the inference both live on your machine. Ollama reduces this to one command: install it, pull a model, and from then on the whole conversation happens on your computer with nothing leaving the machine. The trade-off is model size is capped by your RAM/VRAM, and you download several GB of weights first.
从这里开始
- OllamaMIT · One command
Ollama 是一个给想在本地跑大模型的人用的运行时，能让你一条命令就在自己的电脑上开聊
- Open WebUIOpen WebUI License · Needs tinkering
Open WebUI 是一个给本机的大模型配一个自己的聊天界面，数据不出门，能。
先看哪个
If your machine has little RAM and you want a quick trial, start with a small model in Ollama; if you want to use it in a browser without a CLI, start with the path that offers a web UI.