How to Run Phi-4 on AWS EC2: A Beginner’s Guide

Artificial intelligence models like Phi-4 bring immense power to text generation and processing. In this guide, we will walk through how to set up Phi-4 using Docker and Ollama, ensuring a smooth and isolated environment for AI execution.

Before running Phi-4, ensure that you have access to a Linux based system (Ubuntu recommended) with Docker installed. If you're using an AWS EC2 instance, connect to it first:

ssh -i "your-key.pem" ubuntu@your-ec2-ip

Then, update system packages:

sudo apt update && sudo apt upgrade -y

If Docker isn’t already installed, add it with:

sudo apt install docker.io -y
sudo systemctl start docker
sudo systemctl enable docker

To run Phi-4, you first need Ollama, a containerized environment optimized for large models. Set up Ollama with:

docker run -d --name ollama -v ollama:/root/.ollama -p 11434:11434 ollama/ollama

This initializes a background container with persistent storage.

Once the Ollama container is running, access it using:

docker exec -it ollama /bin/bash

Inside the container, pull Phi-4:

ollama pull phi:4

This step downloads all necessary dependencies.

After installation, start the Phi-4 model with:

ollama run phi:4

To check if it's working, enter a sample query:

>>> What are the key differences between deep learning and machine learning?

To simplify interactions, you can deploy a browser based UI:

docker run -d --name ollama-ui -p 4200:8080 -e 
LLAMA_BASE_URL=http://<YOUR-EC2-IP>:11434 -v open
webui:/app/backend/data --restart always
ghcr.io/open-webui/open webui:main

Access the UI via http://<YOUR-EC2-IP>:4200.

Deploying Phi-4 using Docker and Ollama offers a seamless way to interact with the model while maintaining an efficient environment. Whether using the command-line or a web UI, this setup ensures easy access and optimal performance.

Ready to transform your business with our technology solutions? Contact Us  today to Leverage Our AI/ML Expertise.