🦙 Llama Boilerplate Project

Boilerplate for working with the Llama API via the Ollama platform.

Structure

Ollama-boilerplate/
├── prompts/                        # Folder containing prompt text files
│   ├── init_terminal.txt           # Welcome print content
│   ├── system_prompt.txt           # System prompt content
├── utility/                        # Utility scripts
│   ├── build_messages.py           # Builds messages for the Ollama API
│   ├── call_Ollama.py              # Handles API calls to Ollama
│   ├── check_api.py                # Validates the API key
│   ├── debug_print.py              # Prints debug information
│   ├── display_init_terminal.py    # Initialize the welcome message inside the terminal
│   ├── load_api.py                 # Loads the API key from .env
│   ├── print_response.py           # Print the Ollama reponse with a typewriter effect
│   ├── read_prompt.py              # Reads content from the prompt text file
├── .env                            # Environment file (API endpoint)
├── .gitignore                      # Git ignore file
├── main.py                         # Main script to run the project
├── README.md                       # Project documentation
├── requirements.txt                # List of dependencies

---

Prerequisites

1. Python 3.8+

2. Ollama Installed and Running

   • Visit ollama.com to download and install Ollama.
   • Once installed, ensure the server is running. By default, it should be available at:
     http://localhost:11434/.
   • If the server is not running, use the following commands:

     ollama serve

3. Model Downloaded

   • Models must be downloaded before running the project. For example, to download the llama3.2 model:

     ollama pull llama3.2

   Note: Without a downloaded model, the project will not work. Ensure the model is ready before proceeding.

---

Setting Up the Project

Step 1: Install Dependencies

Install the required Python libraries from requirements.txt:

pip install -r requirements.txt

Step 2: Create the .env File

The .env file is used to store the Ollama API endpoint. You can create it inside the project folder via the command line:

echo "OLLAMA_API=http://localhost:11434/" > .env

Ensure the server URL matches the one provided by Ollama.

Step 3: Fill the prompts/ Directory

Ensure that the prompts/ folder contains the following file:

• system_prompt.txt: A text file with the system prompt.

You can store differen system prompt files in this folder and suit them to your use case.

Step 4: Run the Main Script

Run the main.py script with custom file paths and model names as parameters:

python3 main.py
--system_prompt ./prompts/system_prompt.txt
--model llama3.2

or

python3 main.py
--sp ./prompts/system_prompt.txt
--m llama3.2

---

Troubleshooting

• Server Not Running: Visit http://localhost:11434/. If it's not available:

  ollama serve

• Model Not Found: Download the required model using:

  ollama pull <model-name>

• Dependency Issues: Run pip3 install -r requirements.txt to install all required dependencies.
• API Errors: Check the OLLAMA_API variable in your .env file and ensure it points to a running server.

Your problem is not mentioned? Feel free to ask about it to me or by submitting an issue.

---

License

This project is open-source and available under the MIT License.