Summer School Program on Generative AI Practicals

Core Periods Modules

The Summer School on Generative AI 2025 is structured over five weekends, with sessions held on Saturdays and Sundays in a flexible hybrid format, offering both online participation and offline learning at the Vadodara Learning Centre. The programme comprises ten intensive sessions, carefully designed to blend theoretical understanding with hands-on application. It begins with an introduction to the fundamentals of generative AI and its core architectures, including LLMs, GANs, transformers, and diffusion models. In the following weeks, participants engage in practical work with LLMs and GANs, master prompt engineering techniques, and explore multimodal AI for text, images, audio, and video. The fourth weekend is focused on developing AI agents using Retrieval-Augmented Generation (RAG) and LangChain, empowering participants to build intelligent task-based AI systems. The final weekend delves into advanced language modeling with transformers, fine-tuning models, and culminates in a capstone project where participants apply their learning to real-world challenges. This comprehensive and future-focused programme ensures participants gain the practical skills and confidence needed to leverage generative AI effectively in their careers or ventures.

1. Period 0

   Week 0 – Introduction & Core Architectures

   Overview of Generative AI Understanding Core Architectures: LLMs, GANs, Transformers, and Diffusion Models Evolution and Applications of Generative AI Setting Up AI Tools and Environments

2. Period 1

   Week 1 – Hands-on with LLMs & GANs

   Practical Introduction to Large Language Models (LLMs) Working with GANs for Image Generation Text Generation and Language Understanding Building Simple Generative Models (Text & Images)

3. Period 2

   Week 2 – Multimodal AI & Prompt Engineering Mastery

   Understanding Multimodal AI (Text, Image, Audio, Video) Techniques for Crafting Effective Prompts Prompt Engineering for LLMs, Image Generators, and Multimodal Models Practical Labs with Leading AI Platforms (e.g., OpenAI, MidJourney, DALL·E, Stable Diffusion)

4. Period 3

   Week 3 – Fundamentals of AI Agents Using RAG and LangChain

   Introduction to Retrieval-Augmented Generation (RAG) Building AI Agents with LangChain Designing Task-Oriented AI Agents and Chatbots Integrating External Data Sources and APIs with AI Agents

5. Period 4

   Week 4 – Generative AI Language Modeling with Transformers

   Deep Dive into Transformer Architectures Fine-tuning Pre-trained Models Building Custom Generative Language Applications Future of Generative AI, Capstone Project, and Final Presentation

6. Period 5

   Week 5 - Real-World Applications and Deployment of Generative AI

   To equip participants with practical skills for deploying generative AI models across real-world domains such as business, education, media, design, and healthcare. This module emphasizes integration, scaling, and ethical concerns in production settings.