Ollama 0.1: pull and run local LLMs with one command, now on Linux too
In one sentence On September 23, 2023 Ollama v0.1.0 ships: after the July Mac preview, the tool to download and run LLMs locally with a single command comes to Linux, with NVIDIA GPU acceleration on by default.
Docker made it simple to start any software with one command, without worrying about installing dependencies. Ollama does something similar for AI models: type "ollama run llama2" and, after the download, you have a chatbot answering in the terminal.
The first version (v0.0.1, July 2023) was a preview. With v0.1.0 in September Ollama comes to Linux too and automatically uses the NVIDIA GPU, loading as much of the model as fits.
Simplicity is the key feature: no Python installation, no virtual environment, no manual configuration. Within months it becomes one of the most used tools for local models.
Companies
Ollama
Tools
Ollama
Tags
Sources