How to setup ollama on windows and linuc OSs

fresh code test/code

Markdown

/codeconsole.log('test')[object HTMLPreElement]

This script will:

• Download the ollama binary.
• Create an ollama user and group.
• Set up a systemd service to run Ollama in the background.
• Detect your NVIDIA GPU (if nvidia-smi is working) and configure CUDA support automatically.

Step 3: Verify the Installation

1. Check that the ollama service is active and running:Bashsudo systemctl status ollama You should see active (running). If not, try sudo systemctl start ollama.
2. Check the version (you don't need sudo for this):Bashollama --version

If both commands work, you are ready.

🚀 Using Ollama (The Fun Part)

The commands are identical for Windows and Linux. Open your terminal (PowerShell, CMD, or Linux terminal).

1. Run Your First Model

The ollama run command does two things: it pulls the model if you don't have it, and then it runs it.

Let's start with llama3, the 8-billion parameter instruct model from Meta.

Bash

[object HTMLPreElement]

The first time you run this, you'll see a download progress bar. It's a ~4.7 GB download, so be patient.

Once downloaded, you'll be dropped into a chat prompt:

[object HTMLPreElement]

Start chatting! Ask it to write a Python script, explain a concept, or draft an email.

2. Useful Commands

Here are the other commands you'll use all the time.

• <b>List Your Local Models:</b> See what you have downloaded.Bashollama list This will show you the model name, ID, size, and when it was last modified.
• <b>Pull a Model (Without Running):</b> Maybe you want to download a model to use later.Bashollama pull phi3:mini
• <b>Remove a Model:</b> Free up that disk space.Bashollama rm llama3
• <b>Exit the Chat:</b> Inside a model's chat prompt, just type:/bye

What's Next? The API

The real power of Ollama for a developer is that it automatically serves a REST API at http://localhost:11434.

/codeYou can now use LangChain, Python scripts, or any application to interact with your local models programmatically, for free. You've just become your own AI provider.