Ollama Demo

A demonstration project showcasing local AI model deployment and interaction using Ollama.

Demo Examples

Custom AI agent (Jarvis)

package.py (llama2, prompt = "What is capitalism?")

sample.py (mistral, prompt = "What happened in Korea in 1950?")

Prerequisites

• Python 3.10.11
• Command line interface (Terminal/CMD)
• Internet connection for initial setup

Installation

1. Install Ollama

1. Download the Ollama setup file from https://ollama.com/
2. Run the installer and follow the installation instructions
3. Verify the installation by opening your command line interface and typing:

   ollama

   If Ollama is properly installed, you should see the help information.

2. Install AI Models

Install the required models for this demo:

ollama run mistral
ollama run llama2
ollama run llama3.2

Note: Model downloads may take some time depending on your internet connection, as these are large files.

3. Testing Model Installation

After installation, you can interact with any model directly:

ollama run mistral

• Ask questions and interact with the model locally
• To exit the model interface, type /bye

Project Setup

1. Clone Repository

git clone [repository-url]
cd ollama-demo

2. Create Virtual Environment

# Create virtual environment
python -m venv venv

# Activate virtual environment
# On Windows:
venv\Scripts\activate

# On macOS/Linux:
source venv/bin/activate

3. Install Dependencies

pip install -r requirements.txt

Usage

Running the Demo

With your virtual environment activated, execute the following commands:

# Run the main package
python ../../package.py

# Run sample requests
python ../../sample_requests.py

Creating Custom AI Agents

You can create custom AI agents using Modelfiles:

ollama create [custom-name] -f [Modelfile]

Replace [custom-name] with your desired agent name and [Modelfile] with the path to your Modelfile.

Available Models

This demo includes the following pre-configured models:

• Mistral: A 7B parameter model optimized for instruction following
• Llama2: Meta's foundation model for general-purpose tasks
• Llama3.2: Latest version with improved performance and capabilities

Troubleshooting

Common Issues

1. Ollama command not recognized: Ensure Ollama is properly installed and added to your system PATH
2. Model download fails: Check your internet connection and available disk space
3. Virtual environment issues: Ensure you've activated the virtual environment before running Python scripts

Getting Help

• Visit Ollama Documentation for detailed guides
• Check model-specific documentation for advanced usage

Project Structure

ollama-demo/
├── package.py
├── sample_requests.py
├── requirements.txt
├── README.md
└── venv/

Contributing

Feel free to submit issues and enhancement requests to improve this demo.

License

MIT lincence

---

Note: This demo is designed for local development and testing. Ensure you have sufficient system resources for running large language models locally.