This repository provides a hands-on workshop for developers and solution builders to learn how to leverage foundation models (FMs) through Amazon Bedrock. It covers common Generative AI use cases like text and image generation, retrieval-augmented generation (RAG), model customization, and agents, enabling users to improve productivity by composing emails, summarizing text, and creating images.

How It Works

The workshop is structured as a series of Python notebooks that interact with the Amazon Bedrock API. It demonstrates various techniques for utilizing FMs, including text generation, image generation, and multimodal embeddings. The approach emphasizes practical application through guided labs, with options to integrate with popular open-source frameworks like LangChain and FAISS for advanced RAG and chatbot functionalities.

Quick Start & Requirements

• Install/Run: Clone the repository and run Python notebooks within a chosen environment.
• Prerequisites: An AWS account with IAM permissions for Amazon Bedrock. Recommended environment: SageMaker Studio or SageMaker Notebook Instance. Local environments require AWS credentials.
• Setup: Follow instructions at https://catalog.us-east-1.prod.workshops.aws/amazon-bedrock/en-US.
• Links: Workshop Website

Highlighted Details

• Covers text generation, image generation (Titan, Stable Diffusion XL), and multimodal embeddings.
• Includes labs on managed RAG, LangChain RAG, and model customization (Titan Lite, Llama2 fine-tuning).
• Demonstrates building AI agents for customer service and insurance claims.
• Features optional labs for LangChain examples and NVIDIA NeMo Guardrails.

Maintenance & Community

This is an AWS sample repository, indicating official support and potential for ongoing updates. Community contributions are acknowledged via Star History.

Licensing & Compatibility

The repository is licensed under the Apache License 2.0, permitting commercial use and integration with closed-source projects.

Limitations & Caveats

This workshop must be run within a user's own AWS account and is not compatible with AWS Workshop Studio. Specific AWS region availability for Bedrock models may vary.