SLM Builder Documentation

Welcome to the comprehensive documentation for SLM Builder - an end-to-end Python toolkit for creating, training, and deploying specialized language models.

🚀 Getting Started

New to SLM Builder? Start here:

1. Installation Guide - Get SLM Builder up and running
2. Quick Reference - Cheat sheet for common tasks
3. Examples - Learn by example

📚 Core Documentation

Features

• Core Features - Dynamic model loading, quantization, evaluation
• Additional Features - Database loading, API integration, model comparison

Development

• Developer Guide - Contributing, development setup, release process
• Changelog - Version history and changes

🎯 Quick Navigation

By Use Case

• Building QA Models: Quick Reference → Examples
• Database Integration: Additional Features → Examples
• API Data Loading: Additional Features → Examples
• Model Comparison: Additional Features → Examples

For Developers

• Contributing: Developer Guide - How to contribute
• Development Setup: Developer Guide - Set up your dev environment
• Release Process: Developer Guide - Creating releases
• Testing: Developer Guide - Running and writing tests

By Feature

• Data Sources: Core Features
• Model Loading: Core Features
• Training Methods: Core Features
• Evaluation Metrics: Core Features
• Export Formats: Core Features

🔧 Advanced Topics

• Dataset Splitting & Validation: Additional Features
• Experiment Tracking: Additional Features
• Quantization: Core Features
• Custom Preprocessing: Examples

🤝 Community

• GitHub Repository
• Issue Tracker
• Discussions

📦 Installation

# Basic installation
pip install slm-builder

# With database support
pip install slm-builder[db]

# With all features
pip install slm-builder[all]

See Installation Guide for detailed instructions.

✨ Key Features

• 📥 14 Data Sources: CSV, JSONL, SQL, MongoDB, REST APIs, and more
• 🎯 Multiple Tasks: QA, classification, generation, instruction-tuning
• 🚀 Easy Training: Pre-configured recipes (LoRA, full fine-tuning)
• 💻 CPU & GPU Support: Optimized for both environments
• 📊 Model Comparison: Benchmark multiple models
• 🔬 Advanced Metrics: Perplexity, BLEU, ROUGE, F1
• ⚡ Quantization: 4-bit and 8-bit compression
• 📦 Multiple Exports: ONNX, TorchScript, HuggingFace

🎓 Learning Path

Beginner

1. Read Installation Guide
2. Try Quick Reference examples
3. Explore Basic Examples

Intermediate

1. Learn about Core Features
2. Try Database Loading
3. Experiment with Model Comparison

Advanced

1. Study Additional Features
2. Read Developer Guide
3. Create custom implementations

Contributing

1. Review Developer Guide
2. Set up Development Environment
3. Understand Release Process

📝 Documentation

This documentation is automatically published to GitHub Pages:

• Live Site: https://isathish.github.io/slm/
• Source Files: docs/

Documentation is automatically updated on every push to the main branch.

---

Version: 1.0.0
Last Updated: December 2, 2025
Status: ✅ Production Ready