Skip to content

Latest commit

 

History

History
50 lines (34 loc) · 1.56 KB

README.md

File metadata and controls

50 lines (34 loc) · 1.56 KB
Raw

Movie Summarization Pipeline

This repository contains code for deploying a movie summarization pipeline using AWS services and Python-based Lambda functions. The pipeline utilizes Large Language Models for generating summaries of movies uploaded to an AWS S3 bucket.

Overview

The pipeline is designed to process video files uploaded to S3, using AWS services for event handling, task orchestration, and data storage. It offers both a full and a simplified version for different use cases.

Features

  • AWS S3 Integration: Upload and store movie files.
  • Event-Driven Architecture: Trigger processes with AWS EventBridge and Step Functions.
  • Data Storage: Store and retrieve data with Amazon DynamoDB.
  • API Access: Interact with the pipeline through AWS API Gateway and Lambda functions.
  • LLM Summarization: Leverage Bedrock and Anthropic's Claude v2 for generating summaries.

Getting Started

Prerequisites

  • AWS account and CLI configured.
  • Terraform installed for infrastructure deployment.
  • Python 3.8 for Lambda functions.

Deployment

  1. Initialize Terraform 
    terraform init
  2. Apply Terraform Configuration 
    terraform apply

Usage

  • Upload a video file to the designated S3 bucket.
  • The pipeline will trigger and process the file.
  • Access the summaries through the provided API endpoints.

API Reference

  • List Movies Endpoint: GET /list
  • Summarize Movie Endpoint: GET /summarize/:id

Acknowledgments

  • Anthropic for the Claude v2 model.
  • AWS services used in this project.