FAIR-LLM Installation Guide

🚀 Quick Installation

Prerequisites

• Python 3.8 or higher
• pip package manager

Step 1: Clone the Repository (for demos)

git clone [email protected]:USAFA-AI-Center/fair_llm_demos.git
cd fair-llm-demos

Step 2: Install All Dependencies

Simply install everything needed using the requirements file:

pip install -r requirements.txt

This will install:

• fair-llm>=0.1 - The core FAIR-LLM package
• python-dotenv - For environment variable management
• rich - For beautiful terminal output
• anthropic - For Anthropic Claude integration
• faiss-cpu - For vector search capabilities
• seaborn - For data visualization
• pytest - For testing

Step 3: Set Up API Keys

Create a .env file in your project root:

# Copy the example file
cp .env.example .env

# Or create a new one
echo "OPENAI_API_KEY=your_openai_api_key_here" > .env
echo "ANTHROPIC_API_KEY=your_anthropic_api_key_here" >> .env

Or export them as environment variables:

export OPENAI_API_KEY="your_openai_api_key_here"
export ANTHROPIC_API_KEY="your_anthropic_api_key_here"

Step 4: Verify Installation

Run the verification script:

python verify_setup.py

You should see a colorful output showing all components are properly installed!

🎯 Running the Demos

Once installed, try the demo scripts:

Essay Autograder Demo

# Basic grading
python demos/demo_committee_of_agents_essay_autograder.py \
  --essays essay_autograder_files/essays_to_grade/ \
  --rubric essay_autograder_files/grading_rubric.txt \
  --output essay_autograder_files/graded_essays/

# With RAG fact-checking
python demos/demo_committee_of_agents_essay_autograder.py \
  --essays essay_autograder_files/essays_to_grade/ \
  --rubric essay_autograder_files/grading_rubric.txt \
  --output essay_autograder_files/graded_essays/ \
  --materials essay_autograder_files/course_materials/

Code Autograder Demo

# Static analysis only (safer)
python demos/demo_committee_of_agents_coding_autograder.py \
  --submissions coding_autograder_files/submissions/ \
  --rubric coding_autograder_files/rubric.txt \
  --output coding_autograder_files/reports/ \
  --no-run

# With test execution (requires sandbox)
python demos/demo_committee_of_agents_coding_autograder.py \
  --submissions coding_autograder_files/submissions/ \
  --tests coding_autograder_files/tests/test_calculator.py \
  --rubric coding_autograder_files/rubric.txt \
  --output coding_autograder_files/reports/

📦 Upgrading

To upgrade to the latest versions:

# Upgrade all packages
pip install --upgrade -r requirements.txt

# Or just upgrade fair-llm
pip install --upgrade fair-llm

🐛 Troubleshooting

Missing Dependencies

If you get import errors, ensure all requirements are installed:

pip install -r requirements.txt --force-reinstall

API Key Issues

The demos will create sample files if they don't exist, but ensure your API keys are set:

python -c "import os; print('OpenAI Key:', 'Set' if os.getenv('OPENAI_API_KEY') else 'Not Set')"

Virtual Environment Issues

Always use a virtual environment to avoid conflicts:

python -m venv venv
source venv/bin/activate
pip install -r requirements.txt

📚 What's Included

After installation, you'll have:

• ✅ The complete FAIR-LLM framework
• ✅ Multi-agent orchestration capabilities
• ✅ Document processing tools
• ✅ Vector search with FAISS
• ✅ Beautiful terminal output with Rich
• ✅ Complete demo applications

🎉 Next Steps

1. Run python verify_setup.py to confirm everything is working
2. Explore the demos/ folder for examples
3. Set up and run some demos
4. Start building your own multi-agent demo files!

👥 Contributors

Developed by the USAFA AI Center team:

Ryan R ([email protected]) Austin W ([email protected]) Eli G ([email protected]) Chad M ([email protected])