LLM Comparison Leaderboard
An interactive dashboard for comparing the performance of state-of-the-art large language models across various tasks and metrics.
Features
- Overall model rankings with comprehensive scoring
- Task-specific performance analysis
- Metric breakdowns across different dimensions
- User satisfaction and experience metrics
- Interactive visualizations using Recharts
- Responsive design for all device sizes
Getting Started
Prerequisites
- Node.js 16.8 or later
- Python 3.8 or later (for data processing)
- Python packages: pandas, numpy
Installation
- Clone the repository:
git clone https://github.com/yourusername/llm-comparison-leaderboard.git
cd llm-comparison-leaderboard
- Install dependencies:
npm install
- Install Python dependencies (if you plan to process data):
pip install pandas numpy
Using Sample Data
The repository includes a sample JSON file with placeholder data in public/llm_comparison_data.json
. You can start the development server right away to see the dashboard with this data:
npm run dev
Visit http://localhost:3000 to see the dashboard.
Processing Your Own Data
If you have your own data, follow these steps:
- Place your CSV data file in the
data
directory:
mkdir -p data
cp /path/to/your/pilot_data_n20.csv data/
- Run the data processing script:
npm run process-data
This will:
- Process the CSV data using the Python script
- Generate a JSON file in the
public
directory - Format the data for the dashboard
- Start the development server:
npm run dev
Project Structure
app/
- Next.js App Router componentspage.js
- Main page component that loads data and renders dashboardlayout.js
- Layout component with metadata and global stylesglobals.css
- Global styles including Tailwind CSS
components/
- React componentsLLMComparisonDashboard.jsx
- The main dashboard component
public/
- Static filesllm_comparison_data.json
- Processed data for the dashboard
lib/
- Utility functionsutils.js
- Helper functions for data processing
scripts/
- Data processing scriptsprocess_data.js
- Node.js script for running Python processorprocess_data.py
- Python script for data processing
Building for Production
To build the application for production:
npm run build
To start the production server:
npm run start
License
This project is licensed under the MIT License - see the LICENSE file for details.