Mastering Generative AI: From Basics to Advanced Applications
Journey Through LLMs, Retrieval-Augmented Generation, and Agentic AI
Duration
3 Days (8 hours per day)
Level
Advanced Level
Design and Tailor this course
As per your team needs
Overview
This intensive 3-day course aims to equip participants with a thorough understanding of Generative AI, focusing predominantly on open-source Large Language Models (LLMs). Participants will explore the foundational concepts, practical applications, and advanced techniques in Generative AI, including Retrieval-Augmented Generation (RAG) and Agentic AI, all without relying on cloud deployment. The course combines theoretical insights with hands-on sessions to ensure a comprehensive learning experience.
Audience
- AI enthusiasts and practitioners seeking to deepen their knowledge of Generative AI and open-source LLMs
- Data scientists and machine learning engineers aiming to implement advanced AI solutions without cloud dependencies
- Researchers and academicians interested in the latest developments in Generative AI and its applications
- Technical professionals transitioning into AI roles
Prerequisites
- Basic understanding of Artificial Intelligence or Machine Learning concepts
- Familiarity with programming such as Python
- No prior experience with Generative AI or LLMs is required, as the course will cover topics from beginner to advanced levels
Curriculum
- Overview of Generative AI
-
- What is Generative AI?
- Significance and Use Cases of Generative AI
- Historical context and evolution of systems – Traditional vs ML based vs GenAI vs Agentic AI
- Applications across various industries
- Introduction to Large Language Models (LLMs)
- What is LLM?
- What are Language Models?
- What is “Large” in LLMs?
- Understanding LLM Architecture
- Differences between open-source and proprietary LLMs
- Ethical considerations in LLM deployment
- Exploring Open-Source LLMs
- Survey of popular open-source LLMs (e.g., LLaMA, Falcon, Mistral)
- Understanding Architectures of LLaMA, Falcon, Mistral, etc.
- Installation and setup of open-source LLMs locally
- Hands-on session: Running a simple open-source LLM
- Data Preparation for LLMs
- Sourcing and curating datasets for training LLMs
- Data cleaning and preprocessing techniques
- Hands-on session: Preparing a dataset for LLM training
- Fine-Tuning Open-Source LLMs
- Understanding the fine-tuning process
- Factors for fine-tuning
- Parameter-efficient fine-tuning methods
- Hands-on session: Fine-tuning an open-source LLM for a specific task
- Evaluation and Optimization
- Metrics for assessing LLM performance
- Techniques for optimizing LLM efficiency
- Hands-on session: Evaluating and optimizing a fine-tuned LLM
- Introduction to Retrieval-Augmented Generation (RAG)
- Concept and significance of RAG
- Understanding RAG implementations
- Various Available Tools for RAG Implementation
- Integrating retrieval mechanisms with LLMs
- Hands-on session: Implementing a simple RAG system
- Exploring Agentic AI
- Understanding Agentic AI and its applications
- Current trends and future directions
- Case studies of Agentic AI implementations
- Building Applications with Open-Source LLMs
- Designing AI applications without cloud dependencies
- Tools and frameworks for local deployment
- Hands-on session: Developing a simple AI application using an open-source LLM
- Ethical and Legal Considerations
- Understanding the ethical implications of AI applications
- Compliance with legal standards and regulations
- Best practices for responsible AI development
- Capstone Project
- Participants will work in groups to design and implement a mini-project that incorporates Generative AI, open-source LLMs, Agentic AI and RAG techniques
Duration
3 Days (8 hours per day)
Level
Advanced Level
Design and Tailor this course
As per your team needs
