Mastering LLM Engineering
Dive into the world of Large Language Models (LLMs) in this comprehensive 2-day workshop. This workshop is designed to equip AI engineers, NLP professionals, and developers with the practical skills to build, fine-tune, and deploy state-of-the-art LLMs while adhering to industry best practices in LLMOps.
Learning Objectives
By the end of this workshop, participants will:
- Understand the core principles of LLMs, including architecture, training cycles, and evaluation metrics.
- Gain hands-on experience in supervised fine-tuning and building scalable LLM pipelines.
- Learn how to implement Retrieval-Augmented Generation (RAG) for real-world applications.
- Develop expertise in deploying end-to-end LLM solutions using cloud-based tools like AWS and Azure.
- Master LLMOps practices, including orchestrators, monitoring, and continuous improvement.
Duration: 2 days
Day 1: Foundations and Practical Skills
- Session 1: Introduction to LLM Engineering: Overview of LLM technology and the AI landscape, Core principles of LLM design and implementation.
- Session 2: Data Pipelines and Preprocessing: Building robust data pipelines for LLMs, Hands-on exercise: Preprocessing datasets for model training.
- Session 3: Fine-Tuning LLMs: Techniques for supervised fine-tuning, Hands-on exercise: Fine-tuning an open-source LLM.
- Session 4: LLM Evaluation and Metrics: Evaluating model performance and addressing bias, Hands-on exercise: Using evaluation frameworks.
Day 2: Advanced Applications and Deployment
- Session 1: LLMOps Best Practices, Exploring orchestrators and monitoring tools for LLMs, Case studies: Real-world LLMOps implementations.
- Session 2: Building RAG Applications: Introduction to Retrieval-Augmented Generation, Hands-on exercise: Implementing a RAG pipeline.
- Session 3: Deployment Strategies: Deploying LLM solutions with AWS and other cloud tools, Hands-on exercise: Deploying a fine-tuned LLM.
- Session 4: Final Project and Q&A: Participants implement a complete LLM solution using workshop concepts, Wrap-up and discussion of future trends in LLMs.
