Generative AI

9,900.00 (Inc. GST)

SKU: N/A Category:

Generative AI is a rapidly evolving field that focuses on creating new, synthetic data similar to existing datasets. This course provides a structured approach to understanding fundamental concepts, techniques, and tools required to build and deploy generative models. Participants will explore various generative models, including Generative Adversarial Networks (GANs), Variational Autoencoders (VAEs), Diffusion Models, and Transformers, and gain hands-on experience in applying them to real-world problems.

Pre-Requisites:

  • Basic programming knowledge (preferably Python).
  • Familiarity with Python libraries like NumPy, Pandas, Matplotlib, and TensorFlow/PyTorch.
  • Some understanding of probability distributions and linear algebra.
  • Exposure to deep learning concepts (helpful but not mandatory).

Key Learning Outcomes:

By the end of this course, participants will:

  1. Understand Generative AI and its applications across industries (image generation, text synthesis, data augmentation, etc.).
  2. Differentiate between generative models like GANs, VAEs, Diffusion Models, and Transformers.
  3. Gain hands-on experience in implementing and fine-tuning generative models using TensorFlow/PyTorch.
  4. Learn evaluation metrics like Frechet Inception Distance (FID), Inception Score (IS), and Perplexity.
  5. Explore ethical considerations, including bias, misinformation, adversarial attacks, and copyright issues.
  6. Deploy generative models using Hugging Face, Streamlit, or OpenAI API.

Target Audience:

  • AI enthusiasts and professionals looking to expand their knowledge of Generative AI.
  • Students and professionals from non-technical backgrounds who want to understand AI-generated content.
  • Developers and data scientists interested in building real-world applications using Generative AI.

Test & Evaluation:

  • Assignments: Regular hands-on tasks for skill reinforcement.
  • Final Project: Developing and presenting a generative AI model.
  • Assessment: MCQs + practical evaluation.
  • Certification:
    • Certificate of Completion upon successful course completion.
    • Project Letter for those who complete the final project.
    • No certification for participants who leave the course midway.

Delivery Mode & Duration:

  • Mode: Online Live Sessions
  • Total Duration: 120 Hours (60 Hours Live Sessions + 60 Hours Hands-on Assignments)

Additional information

Centre for Summer Training

IIT Kanpur Campus, Online Live

Batch Date

Batch 1: 19th May 2025 – 25th June 2025, Batch 2: 17th June 2025 – 22nd July 2025

Curriculum

Module 1: Introduction to Generative AI

  • What is Generative AI?
  • Historical Context and Evolution of AI
  • Applications in healthcare, finance, creative arts, and gaming
  • Overview of Generative Models: GANs, VAEs, Diffusion Models, Transformers
  • Introduction to Prompt Engineering
  • Ethical Concerns: Bias, Copyright Issues, Adversarial Attacks

Module 2: Core Concepts in AI & Deep Learning

  • Basics of Machine Learning and AI
  • Introduction to Neural Networks
  • Backpropagation and Gradient Descent
  • Overview of Probability Distributions in AI
  • Introduction to Transformers and Attention Mechanisms

Module 3: Generative Adversarial Networks (GANs)

  • Understanding GANs: Generator vs. Discriminator
  • How GANs Work: Training, Loss Functions, Optimization
  • Hands-on: Implementing a Basic GAN in TensorFlow/PyTorch
  • Advanced GAN Variants: DCGAN, WGAN, StyleGAN
  • Applications: Deepfakes, Super-Resolution, Data Augmentation
  • Troubleshooting mode collapse, vanishing gradients, and training instability

Module 4: Variational Autoencoders (VAEs)

  • Introduction to Autoencoders
  • How VAEs differ from GANs
  • Mathematical Foundation: Latent Space Representation
  • Hands-on: Implementing VAEs in Python
  • Applications: Data Compression, Image Generation, Anomaly Detection

Module 5: Diffusion Models & Advanced Generative Techniques

  • Introduction to Diffusion Models
  • Comparison: GANs vs. VAEs vs. Diffusion Models
  • Hands-on: Implementing a Simple Diffusion Model
  • Introduction to Stable Diffusion & DALL-E
  • Text-to-Image Generation: Building a Simple Text-to-Image Model

Module 6: Generative AI in Natural Language Processing (NLP)

  • Introduction to Large Language Models (LLMs)
  • Exploring GPT, BERT, LLaMA, and Claude
  • Fine-Tuning Pretrained LLMs for Domain-Specific Tasks (Healthcare, Finance, etc.)
  • Hands-on: Building AI-powered Text Summarization & Translation
  • Introduction to Chatbot Development with OpenAI API & LangChain

Module 7: Fine-Tuning, Evaluation, and Deployment

  • Fine-Tuning Pretrained Generative Models
  • Metrics for Generative AI: FID, Inception Score, BLEU, Perplexity
  • Hands-on: Fine-tuning and evaluating a GAN or VAE
  • Optimization Techniques to improve model efficiency
  • Deploying Models on Hugging Face & Streamlit

Module 8: Final Project & Future Trends in Generative AI

  • Final Project:
    • Choose between Image Generation (GANs/Diffusion) or Text Generation (GPT/VAEs)
    • Deploying the project on Hugging Face or Streamlit
  • Future of Generative AI: Multimodal AI, Explainable AI, and AI in Creativity
  • Course Wrap-Up & Q&A

Enquiry