API Driven Cloud Native Solutions

Lesson 8 Study Notes

Topic List

• Understanding AI, ML, Deep Learning, and Generative AI
• 2. AI Use Cases
• 3. The Three Generations of Machine Learning
• 4. Pre-trained Models
• 5. Fine-tuned Models
• 6. Transfer Learning
• 7. Pre-trained vs Fine-tuned Models
• 8. Where These Models Live and How They Are Used
• 9. Interpretability in Fine-tuned Models
• 10. Cognitive Services
• 11. Azure Cognitive Services
• 12. Amazon AI Services
• 13. Google Cloud Vertex AI (Gemini Models)
• 14. Categories of Cognitive Services
• 15. Benefits of Using Cognitive Services
• 16. Cognitive Services in the Broader AI Workflow
• 17. Hugging Face
• 18. Tasks in Hugging Face
• 19. Models in Hugging Face
• 20. Datasets in Hugging Face
• 21. How Tasks, Models, and Datasets Interact
• 22. Example Workflow
• 23. Model Cards
• 24. Basics of Natural Language Processing (NLP)
• 25. Understanding Language
• 26. Syntactic Processing
• 27. Semantic Processing
• 28. Natural Language Understanding (NLU)
• 29. Natural Language Generation (NLG)
• 30. Example NLP Pipeline
• 31. Tools for NLP and NLU
• 32. NLP Components and Pipelines
• 33. Transformer Applications in NLP
• 34. Applications of Transformers (NLP)
• 35. Computer Vision APIs
• 36. Key Computer Vision Tasks
• 37. Example Computer Vision API

Lesson 9 Study Notes

Topic List

• Language model fundamentals
• Large vs small models
• Model parameters and training
• Common Crawl
• Tokenisation
• Popular LLMs and SLMs
• LLMOps and the development-to-production workflow

Lesson 10 & 11

• OpenAI APIs Overview
• Access Requirements for Using APIs
• Base Code Setup for OpenAI API
• Chat or Completion API
• Embedding API
• DALL-E API
• Whisper API
• Fine-tuning API
• Rivet
• Retrieval Augmented Generation (RAG)
• LangChain and RAG Implementation Workflow
• RAG Demo and Gradio Interface

Lesson 12

• Docker
• Container

• Kubernetes

Lesson 13

Data Analytics

Lesson 14

Introduction to IoT

Distributed Computing

Lesson 7: Distributed Mutual Exclusion

Topic List

• Introduction to Distributed Mutual Exclusion
• Mutual Exclusion: Fundamental Idea
• Critical Section (CS)
• Preliminaries for Distributed Mutual Exclusion
• System Model
• States of a Site
• Additional State in Token-Based Algorithms
• Requirements of Mutual Exclusion Algorithms
• Performance Metrics
• Approaches to Distributed Mutual Exclusion
• Lamport’s Algorithm
• Ricart–Agrawala Algorithm
• Maekawa’s Algorithm

Deadlock Detection (Distributed Computing)

Lesson 8

Topic List

• Suzuki-Kasami Broadcast Algorithm
• Raymond’s Tree-Based Algorithm
• Deadlock Detection in Distributed Systems
• Wait-For Graph (WFG)
• Models of Deadlocks
• Single-Resource Model
• AND Model
• OR Model
• Unrestricted Model
• Chandy–Misra–Haas Algorithm for AND Model
• Chandy–Misra–Haas Algorithm for OR Model

Lesson 9

Topic List

1. Need for agreement
2. Byzantine behaviour
3. Agreement in failure-free systems
4. Crash-failure consensus algorithm
5. Lower bound on rounds
6. Upper bound on Byzantine failures
7. Byzantine Agreement Tree Algorithm
8. Explanation of example
9. Applying Byzantine Agreement to an 8-node system
10. Consensus and Agreement Algorithms
11. Byzantine Behaviour and Agreement Problem
12. Impossibility of Consensus (FLP Result)
13. Terminating Reliable Broadcast
14. Distributed Transaction Commit (2PC and 3PC)
15. k-Set Consensus

Lesson 10

• Characteristics of P2P Networks
• P2P vs Client–Server Networks
• Churn in P2P Systems
• DNS Limitations and P2P Advantages
• Examples of Popular P2P Networks
• Self-Organisation and Scalability
• Advantages of P2P Overlays
• Characteristics and Performance Features of P2P Systems
• Napster Architecture
• Central Server Metadata in Napster
• Napster Search and Download Process
• Anonymity and Directory-Based Lookup
• Data Indexing in P2P (Centralised, Local, Distributed)
• Structured Overlays
• Unstructured Overlays
• Advantages and Disadvantages of Unstructured Overlays
• Iterative Prisoner’s Dilemma in P2P Design
• Tit-for-Tat Strategy and BitTorrent
• Trust / Reputation Management in P2P
• Challenges in Trust Management Protocol Design

Lesson 14

• CAP Theorem

Comprehensive Question Paper Q&A