Module 1

By the end of this module, learners should be able to:

• Describe what a transformer model is in simple terms

• Explain what tokens and embeddings are

• Explain why attention is important in modern models

• Distinguish between training, fine-tuning, and inference

• Describe what RAG is and when to use it
  
  

Topics

• Transformers and attention

• Tokenization and embeddings

• Fine-tuning, adapters, and prompting

• Inference versus training

• Retrieval-Augmented Generation (RAG)
  
  

Tools Used

• OpenAI Playground or similar

• Gemini Playground or similar

• A transformer visualizer (for example, Neet-style tools)

• HuggingFace Spaces demo apps

Suggested Session Plan (60–90 minutes)

• 10 minutes – Warm up: ask learners to describe how they think AI “reads” text

• 20 minutes – Concept intro with diagrams (transformers, tokens, embeddings, attention)

• 20 minutes – Live demos using playgrounds and visualizers

• 20 minutes – RAG concepts and examples

• 10–20 minutes – Q&A and recap

Hands-On Activities

• Inspect tokenization of short and long sentences

• Use a visual tool to show attention focusing on different words

• Demonstrate a simple RAG example using a hosted tool (upload a document and ask questions about it)