DeAI Bootcamp

Presented by Encode Club

Introduction

Welcome to the Decentralized AI Applications Bootcamp. This repository houses all the necessary instructions and resources for an immersive 8-week program designed to equip students with cutting-edge techniques for developing applications using Generative Artificial Intelligence and Decentralized Computing technologies.

Our bootcamp emphasizes a hands-on and project-based approach, allowing students to learn concepts and immediately apply them to real-world challenges. While we don't dive into the depths of AI and Machine Learning research itself, our focus is on the practical applications of AI technology across various industries.

Throughout this program, students will gain proficiency in utilizing the latest generative AI models and tools. They'll learn to construct applications capable of comprehending natural language inputs and generating diverse content, including text and images.

Upon successful completion of the bootcamp, participants will have acquired:

• A comprehensive overview of state-of-the-art AI models and tools
• Expertise in Web3 technologies for Decentralized Computing for AI
• Familiarity with industry best practices and advanced techniques in AI development
• The capability to independently develop applications incorporating AI features

Curriculum

Our bootcamp is structured as follows:

• Six weeks of intensive, hands-on lessons (Monday to Thursday, four lessons per week) complemented by weekend projects
• A special week featuring four days of guest lectures and interactive workshops
• A final week dedicated to project development and presentation

Week 1: Foundations

• Introduction to AI and Decentralized AI
• Applications of Web3 technologies for AI
• Fundamentals of Machine Learning
• Leveraging Google Colab Notebooks for AI model testing
• Exploring transformers
• Introduction to the Hugging Face Transformers library
• Executing simple AI tasks with the transformers library
• Introduction to GPTs
• Demystifying LLMs
• Using OpenAI APIs
• Configuring a Python development environment
• Understanding model configurations
• Fundamentals of prompt engineering
• Building a basic chatbot using OpenAI APIs

Week 2: Frontend Development and AI Integration

• Setting up a JavaScript/TypeScript development environment
• Introduction to frontend development with React
• Building an interactive chat application with React and Vercel SDK
• Implementing application for streaming chat responses from OpenAI API
• Utilizing sample projects effectively
• Deploying projects on Vercel
• Implementing text-to-image API endpoints
• Implementing text-to-audio API endpoints
• Leveraging AI tools for code generation
• Introduction to Computer Vision
• Implementing a simple multimodal AI chatbot

Week 3: Local AI Model Deployment

• Running Large Language Models (LLMs) on personal hardware
• Cloud services for LLM execution
• Running AI models with Model Loaders
• Exploring open-source AI models
• Setting up local AI inference
• Optimizing configurations for specific hardware
• Serving AI models through a local API
• Replacing OpenAI APIs with a compatible local alternative
• Exploring different open-access AI models
• Mastering advanced prompt techniques

Week 4: Advanced AI Techniques

• In-depth explanation of GPTs
• Model training overview
• Introduction to AI models fine-tuning
• Understanding Low-Rank Adaptations (LoRAs)
• Creating and implementing LoRAs
• Extending GPT capabilities
• AI Assistants overview
• Introduction to Retrieval Augmented Generation (RAG)
• Implementing RAGs with OpenAI Assistants
• Exploring Agentic RAG
• Developing a RAG-based chatbot using custom data

Week 5: Introduction to Decentralized AI

• Introduction to Decentralized Computing and Blockchain Basics
• Introduction to the EVM
• Smart contracts and decentralized computing
• Data structures for blockchain and Oracles
• Verifiable computation and decentralized file storage
• Introduction to Smart Contracts for AI Systems
• Exploring Verifiable AI and Transparency in Model Outputs
• Implementing AI Applications with Blockchain-Backed Data Verification
• Introduction to IMOs and verifiable AI Outputs
• Exploring AI Governance and Token Economic Incentives
• Introduction to Verifiable Credentials in AI Applications
• Exploring Web3 for Data Storage

Week 6: Building Decentralized AI Applications

• Introduction to Function Calling and AI Agents
• Implementing AI Agents
• AI Agent frameworks
• Building Blockchain-Enabled Autonomous AI Agents
• Developing AI Agents with Tokenized Incentive Structures
• Implementing custom function calling capabilities to AI agents
• Deploying Onchain AI Agents on decentralized hosting providers
• Overview of Decentralized Marketplaces for AI Models and Services
• Exploring Federated Learning for Decentralized AI

Bootcamp Material

All bootcamp materials are accessible within this repository. The content is organized into folders, with each folder representing a day's lesson. These folders contain comprehensive lesson materials, including hands-on exercises, implementation examples, illustrated instructions, and step-by-step guides to follow during the lesson.

Important: Each folder contains a README.md file with detailed instructions for the day. Students are strongly advised to clone the repository and prepare all required software and tools prior to the lesson time. This preparation enables students to follow the lessons and complete exercises concurrently with the instructor.

Each lesson requires specific software and tools to be installed on the student's device/computer. The README.md file for each day provides a comprehensive list of required software and tools. The installation and configuration steps may vary in duration, ranging from several minutes to hours, depending on the student's device specifications.

Hands-on Lessons

To maximize learning, students should prepare their own development environment to actively participate in the lessons. Each day, they will learn by replicating the instructor's steps and applying newly acquired concepts to real-world projects.

Note: All lessons will be recorded and made available for later review, allowing students to revisit the material at their own pace if needed.

The recorded material is strictly confidential and must not be shared with anyone outside the bootcamp.

Students are encouraged to ask questions and seek assistance from instructors and mentors during the lessons. Additional support is available through dedicated channels on the bootcamp's Discord server.

We strongly encourage students to collaborate on group projects, share discoveries, learn from peers, and assist one another in completing project challenges.