JuDDGES Tutorials¶
Welcome to the JuDDGES tutorial collection! These hands-on, interactive tutorials will guide you from beginner to expert in legal document analysis with AI.
π Tutorial Overview¶
All tutorials follow the DiΓ‘taxis framework for learning-oriented documentation. Each tutorial:
- β Teaches by doing - Hands-on exercises with real code
- β Builds progressively - Each tutorial builds on previous ones
- β Includes checkpoints - Verify your understanding as you go
- β Provides solutions - Complete working code examples
- β Safe to experiment - Learn by making mistakes
π― Learning Path¶
Follow this recommended path to master JuDDGES:
graph TD
A[Tutorial 1<br/>First Document Analysis<br/>30-60 min] --> B[Tutorial 2<br/>Embeddings<br/>45 min]
B --> C[Tutorial 3<br/>Fine-tuning<br/>60 min]
B --> D[Tutorial 4<br/>Advanced Extraction<br/>45 min]
C --> E[Tutorial 5<br/>End-to-End Project<br/>90 min]
D --> E
style A fill:#e1f5ff
style B fill:#fff3e0
style C fill:#f3e5f5
style D fill:#e8f5e9
style E fill:#fce4ec
Main Tutorial Series¶
Tutorial 1: Your First Legal Document Analysis¶
Learn the fundamentals of legal document analysis with JuDDGES.
What You'll Learn: Set up JuDDGES β’ Load legal datasets β’ Extract information β’ Semantic search β’ Visualizations
Level: π’ Beginner | Duration: 30-60 min | GPU: Not required
Tutorial 2: Working with Legal Document Embeddings¶
Master document embeddings and vector search for legal documents.
What You'll Learn: Generate embeddings β’ Set up Weaviate β’ Ingest documents β’ Semantic search β’ UMAP visualization
Level: π‘ Intermediate | Duration: 45 min | GPU: Optional
Tutorial 3: Fine-tuning Your First Legal LLM¶
Learn to fine-tune large language models for legal tasks.
What You'll Learn: Instruction datasets β’ PEFT/LoRA β’ Training β’ Evaluation β’ Deployment
Level: π΄ Advanced | Duration: 60+ min | GPU: Required (40GB+)
Tutorial 4: Advanced Information Extraction¶
Master advanced extraction techniques with Gemini and LangChain.
What You'll Learn: Complex schemas β’ Multi-step pipelines β’ Validation β’ Scale processing β’ Production deployment
Level: π΄ Advanced | Duration: 45 min | GPU: Not required
Tutorial 5: Building an End-to-End Legal Analysis System¶
Build a complete production-ready legal document analysis pipeline.
What You'll Learn: System design β’ Data pipelines β’ API services β’ Monitoring β’ Deployment β’ Optimization
Level: β« Expert | Duration: 90 min | GPU: Optional
Supplementary Tutorials¶
Setup & Configuration¶
- Getting Started - Quick 30-minute introduction to JuDDGES
- Git LFS Setup - Configure Git Large File Storage for datasets
- Langfuse Setup - Set up LLM observability and monitoring
Feature-Specific¶
- Gemini Extraction - Detailed guide to Gemini API for information extraction
π Tutorial Matrix¶
| Tutorial | Duration | Level | GPU | Focus |
|---|---|---|---|---|
| Tutorial 1 | 30-60 min | π’ Beginner | No | Basics |
| Tutorial 2 | 45 min | π‘ Intermediate | Optional | Embeddings |
| Tutorial 3 | 60+ min | π΄ Advanced | Yes | Fine-tuning |
| Tutorial 4 | 45 min | π΄ Advanced | No | Extraction |
| Tutorial 5 | 90 min | β« Expert | Optional | Production |
Total Learning Time: 4-6 hours for all main tutorials
π By Use Case¶
For Researchers¶
For Data Scientists¶
For Legal Tech Developers¶
For ML Engineers¶
- Skim Tutorial 1
- Deep dive Tutorial 3: Fine-tuning
- Tutorial 5: End-to-End Project
π οΈ How to Use These Tutorials¶
Before You Start¶
- Set up environment: Complete Getting Started Guide
- Check prerequisites: Each tutorial lists required knowledge/tools
- Allocate time: Set aside uninterrupted time
- Prepare workspace: Terminal, editor, browser ready
During the Tutorial¶
- Read first: Understand before coding
- Type yourself: Don't copy-paste
- Complete checkpoints: Verify understanding
- Experiment: Try variations
- Take notes: Document insights
After the Tutorial¶
- Complete exercises: Test knowledge
- Try challenges: Push further
- Build projects: Apply learnings
- Share feedback: Improve tutorials
π‘ Tips for Success¶
Learn by Doing¶
Type code yourself to:
- Understand syntax
- Debug errors
- Build muscle memory
- Gain confidence
Embrace Mistakes¶
Errors are learning opportunities:
- Read error messages carefully
- Check Troubleshooting sections
- Search documentation
- Ask for help
Experiment Freely¶
After each section:
- Modify parameters
- Try different inputs
- Test edge cases
- Break and fix things
π Getting Help¶
Documentation¶
- How-To Guides - Solve specific problems
- Reference - Technical details
- Explanation - Understand concepts
Community¶
- GitHub Issues - Bugs/features
- GitHub Discussions - Questions/ideas
- Email: lukasz.augustyniak@pwr.edu.pl
Common Issues¶
π Track Your Progress¶
- Completed Tutorial 1: First Document Analysis
- Completed Tutorial 2: Embeddings
- Completed Tutorial 3: Fine-tuning
- Completed Tutorial 4: Advanced Extraction
- Completed Tutorial 5: End-to-End Project
- Built a personal project
- Contributed to JuDDGES
- Shared work with community
π€ Contributing¶
Help improve tutorials:
- Report issues: Open an issue
- Suggest improvements: Start a discussion
- Submit changes: Fork, edit, PR
π Start Learning¶
Ready? Begin with:
Tutorial 1: Your First Legal Document Analysis β
Or jump to:
- Tutorial 2: Embeddings
- Tutorial 3: Fine-tuning
- Tutorial 4: Advanced Extraction
- Tutorial 5: End-to-End Project
Last Updated: 2025-10-11 | Version: 1.0 | Status: Published