🚀 AWS Academy - Goodbye World Final Project

Deploy SageMaker Endpoints & Bedrock Runtime Exploration

A comprehensive project for the Great AI Hackathon Malaysia 2025, demonstrating AWS SageMaker endpoint deployment and Amazon Bedrock model integration.

© 2025 Goodbye World team, for Great AI Hackathon Malaysia 2025 usage.

---

📁 Project Structure

🤖 sagemaker/

SageMaker Model Deployment & Notebooks

• Jupyter notebooks for deploying and working with Hugging Face models
• genai-llm.ipynb - Generative AI Large Language Model experiments
• qna-llm.ipynb - Question & Answer system using LLM models
• Demonstrates deployment of models like meta.llama3-8b-instruct-v1:0 and distilbert-base-uncased-distilled-squad for text generation and embeddings

🧠 bedrock/

Amazon Bedrock Foundation Models

• Core Bedrock service integration and setup utilities
• bedrock_wrapper.py - Wrapper class for Bedrock operations
• bedrock_studio_bootstrapper.py - Automated setup for Bedrock Studio environments
• hello_bedrock.py - Basic Bedrock API introduction
• Foundation model management and configuration tools

⚡ bedrock-runtime/

Bedrock Runtime API Examples & Model Comparison

• Extensive collection of model-specific runtime examples
• Supported Models:
  • 🔥 Amazon Nova (Text, Canvas, Reel)
  • 🏛️ Amazon Titan (Text, Image, Embeddings)
  • 🧬 Anthropic Claude (Various versions)
  • 🎯 Cohere Command (R & Standard)
  • 🦙 Meta LLaMA (3.1 variants)
  • 🌪️ Mistral AI
  • 🎨 Stability AI (Image generation)

Special Features:

• 📊 comparison/ - Model comparison tools for testing different AI models with the same prompts
• 🔄 cross-model-scenarios/ - Advanced scenarios like tool use demonstrations
• 🧪 test/ - Comprehensive test suite for all model integrations

---

🛠️ Key Features

🎯 Model Comparison System

Located in bedrock-runtime/comparison/, this system allows you to:

• Compare responses from multiple AI models using identical prompts
• Measure performance and response times
• Export results in various formats
• Interactive and command-line interfaces available

📚 Comprehensive Examples

• Hello World programs for each AI service
• Streaming responses for real-time applications
• Document understanding capabilities
• Image generation workflows
• Cross-model tool usage scenarios

🧪 Testing Infrastructure

• Automated integration tests for all models
• pytest-based testing framework
• Performance benchmarking tools
• Error handling and validation

---

🚀 Quick Start

Prerequisites

# Install Python dependencies
pip install -r bedrock/requirements.txt
pip install -r bedrock-runtime/requirements.txt

# Configure AWS credentials
aws configure

Basic Usage

1. Explore Bedrock Models:

cd bedrock-runtime
python hello/hello_bedrock_runtime_converse.py

2. Compare AI Models:

cd bedrock-runtime/comparison
python quick_compare.py

3. SageMaker Experiments:

cd sagemaker
# Open genai-llm.ipynb in Jupyter
jupyter lab genai-llm.ipynb

---

📖 Documentation

Each folder contains detailed README files with specific instructions:

• bedrock-runtime/comparison/README.md - Model comparison tools
• bedrock-runtime/cross-model-scenarios/tool_use_demo/README.md - Advanced scenarios
• Individual model folders contain usage examples and API documentation

---