Large Language Models Tutorial

Overview

Learn how to create and serve Large Language Models (LLMs) using Digitalhub. This tutorial demonstrates working with HuggingFace-compatible models, including direct model serving from the HuggingFace catalog and fine-tuned model deployment.

What You'll Learn

• Setting up LLM projects in Digitalhub
• Working with HuggingFace model catalog
• Serving pre-trained LLM models
• Deploying fine-tuned models from storage (e.g., S3)
• GPU resource management for LLM workflows
• Using cluster-defined profiles for LLM serving

Getting Started

1. Import the Jupyter notebook notebook-llm-model.ipynb into your Coder instance
2. Execute each cell step by step following the instructions

Resources

• 📁 Source Files: digitalhub-tutorials/s5-llm
• 🤗 HuggingFace Documentation: Transformers library docs