WhiteLightning Model Library

Welcome to the WhiteLightning Model Library - a centralized repository for uploading, downloading, and managing trained machine learning models.

📋 Overview

This repository serves as a comprehensive model library where you can:

• Upload your trained machine learning models
• Download pre-trained models for your projects
• Share models with team members and collaborators
• Access model metadata, training data, and configuration files

🏗️ Repository Structure

The repository is organized by model classification types:

whitelightning-model-library/
├── Binary/                    # Binary classification models
│   ├── hate_speech_classifier/
│   ├── sentiment_classifier/
│   └── spam_classifier/
├── Multiclass/               # Multiclass classification models
│   ├── customer_review_topic/
│   └── news_topic_clf/
└── Multiclass(Sigmoid)/      # Multilabel classification models
    ├── emotion_classifier/
    └── news_multilabel/

🤖 Available Models

Binary Classification Models

1. Hate Speech Classifier

• Purpose: Detects hate speech in text content
• Type: Binary classification (hate speech / not hate speech)
• Files: Model weights (.h5, .onnx), training data, vocabulary, scaler

2. Sentiment Classifier

• Purpose: Analyzes sentiment in text
• Type: Binary classification (positive / negative sentiment)
• Files: Model weights (.h5, .onnx), training data, vocabulary, scaler

3. Spam Classifier

• Purpose: Identifies spam messages
• Type: Binary classification (spam / not spam)
• Files: Model weights (.h5, .onnx), training data, vocabulary, scaler

Multiclass Classification Models

1. Customer Review Topic Classifier

• Purpose: Categorizes customer reviews by topic
• Type: Multiclass classification
• Categories: Customer Service, Ease of Use, Price, and more
• Files: Model weights (.h5, .onnx), training data, vocabulary, scaler

2. News Topic Classifier

• Purpose: Classifies news articles by topic
• Type: Multiclass classification
• Categories: Business, Health, Politics, and more
• Files: Model weights (.h5, .onnx), training data, vocabulary, scaler

Multilabel Classification Models (Sigmoid)

1. Emotion Classifier

• Purpose: Detects multiple emotions in text simultaneously
• Type: Multilabel classification with sigmoid activation
• Categories: Fear, Happy, Love, Sadness
• Unique Feature: Single text can belong to multiple emotion categories
• Files: Model weights (.keras, .onnx), training data, vocabulary, scaler

2. News Multilabel Classifier

• Purpose: Categorizes news articles into multiple topics simultaneously
• Type: Multilabel classification with sigmoid activation
• Categories: Business, Environment, Health, Politics, Sports, Technologies, Wars
• Unique Feature: Single article can belong to multiple news categories
• Files: Model weights (.keras, .onnx), training data, vocabulary, scaler

📁 Model Directory Structure

Each model directory contains:

model_name/
├── model.h5                    # Keras model file
├── model.onnx                  # ONNX model file for cross-platform deployment
├── training_data.csv           # Training dataset
├── edge_case_data.csv          # Edge case test data
├── vocab.json                  # Vocabulary mapping
├── scaler.json                 # Feature scaling parameters
├── generation_config.json      # Model generation configuration
├── *_edge_case_predictions.csv # Edge case prediction results
└── api_requests/               # API request logs and examples
    └── *.json                  # Request/response examples

🚀 How to Use

Downloading Models

1. Clone the repository:

   git clone <repository-url>
   cd whitelightning-model-library

2. Navigate to desired model:

   cd Binary/sentiment_classifier/  # Example

3. Load the model in your code:

   # For Keras models
   from tensorflow.keras.models import load_model
   model = load_model('model.h5')
   
   # For ONNX models
   import onnxruntime as ort
   session = ort.InferenceSession('model.onnx')

Uploading New Models

1. Create a new directory under the appropriate classification type:

   mkdir Binary/your_model_name/
   # or
   mkdir Multiclass/your_model_name/
   # or
   mkdir "Multiclass(Sigmoid)/your_model_name/"

2. Add required files:

   • model.h5 - Keras model
   • model.onnx - ONNX model (optional but recommended)
   • training_data.csv - Training dataset
   • vocab.json - Vocabulary file
   • scaler.json - Scaler parameters
   • generation_config.json - Model configuration
   • edge_case_data.csv - Test cases (optional)

3. Commit and push:

   git add .
   git commit -m "Add new model: your_model_name"
   git push origin main

📊 Model Formats

• Keras (.h5): Native TensorFlow/Keras format
• ONNX (.onnx): Cross-platform format for deployment
• JSON files: Configuration and vocabulary data
• CSV files: Training and test datasets

🔧 Requirements

• Python 3.7+
• TensorFlow/Keras (for .h5 models)
• ONNX Runtime (for .onnx models)
• pandas, numpy (for data handling)

📝 Contributing

1. Fork the repository
2. Create a feature branch
3. Add your model following the directory structure
4. Submit a pull request with model description

⚡ Main Repo

WhiteLightning distills massive, state-of-the-art language models into lightweight, hyper-efficient text classifiers. It's a command-line tool that lets you create specialized models that run anywhere—from the cloud to the edge—using the universal ONNX format for maximum compatibility.

🌐 Documentation & Website

Need comprehensive guides and documentation? Check out our WhiteLightning Site - this repository hosts the official website for WhiteLightning at https://whitelightning.ai, a cutting-edge LLM distillation tool with detailed documentation, tutorials, and implementation guides.

🧪 Testing & Validation

Want to test your ONNX models across multiple programming languages? Check out our WhiteLightning Test Framework - a comprehensive cross-language testing suite that validates your models in:

• 8 Programming Languages: Python, Java, C++, C, Node.js, Rust, Dart, and Swift
• Performance Benchmarking: Detailed timing, memory usage, and throughput analysis
• Automated Testing: GitHub Actions workflows for continuous validation
• Real-world Scenarios: Test with custom inputs and edge cases

Perfect for ensuring your WhiteLightning models work consistently across all target platforms and deployment environments.

📞 Support

For questions or issue regarding the model library, please:

• Open an issues in this repository
• Contact the WhiteLightning team

License

This project is licensed under the MIT License.

Happy modeling! 🚀