Tutorial Notebooks¶
Complete tutorial series for llcuda v2.2.0 on Kaggle dual T4 - 11 comprehensive tutorials from beginner to advanced.
Core Tutorials (1-10)¶
| # | Notebook | Open in Kaggle | Description | Time |
|---|---|---|---|---|
| 01 | Quick Start | 5-minute introduction | 5 min | |
| 02 | Server Setup | Server configuration | 15 min | |
| 03 | Multi-GPU | Dual T4 tensor-split | 20 min | |
| 04 | GGUF Quantization | K-quants, I-quants | 20 min | |
| 05 | Unsloth Integration | Fine-tune → Deploy | 30 min | |
| 06 | Split-GPU + Graphistry | LLM + Visualization | 30 min | |
| 07 | OpenAI API | OpenAI SDK | 15 min | |
| 08 | NCCL + PyTorch | Distributed PyTorch | 25 min | |
| 09 | Large Models (70B) | 70B on dual T4 | 30 min | |
| 10 | Complete Workflow | End-to-end | 45 min |
⭐ Advanced Visualization (Tutorial 11) - MOST IMPORTANT¶
| # | Notebook | Open in Kaggle | Description | Time |
|---|---|---|---|---|
| 11 | GGUF Neural Network Visualization ⭐ | Complete model architecture as interactive graphs | 60 min |
Why Tutorial 11 is Critical: - 🏆 First-of-its-kind: Only comprehensive GGUF visualization tool - 📊 929 nodes, 981 edges: Complete Llama-3.2-3B architecture - 🎨 Interactive dashboards: 8 Graphistry cloud visualizations - 🔬 Research-grade: PageRank, centrality, community detection - 🖥️ Split-GPU showcase: LLM (GPU 0) + Analytics (GPU 1) - 📥 Downloadable: HTML dashboards for offline viewing
What You'll Visualize: - Complete 28-layer transformer architecture - 896 attention heads across all layers - Layer-by-layer breakdowns (35 nodes each) - Q4_K_M quantization block structure - Information flow through the network
Learning Paths¶
Beginner (1 hour)¶
Start here if you're new to llcuda:
Intermediate (3 hours)¶
Full fundamentals with deployment:
Advanced (2 hours)¶
Multi-GPU focus for large models:
Visualization & Research (2.5 hours) ⭐ RECOMMENDED¶
Complete architecture analysis with Graphistry:
Complete Master (6 hours)¶
Everything from basics to advanced visualization: