One-Day Tour Planning Application

Overview

This is a one-day tour planning assistant that helps users create a comprehensive plan for exploring a city based on their preferences. The system uses a combination of FastAPI, Neo4j, and Streamlit to provide a seamless and personalized experience.

Installation

1. Clone the repository
2. Navigate to the project directory
3. Install the required dependencies:

Running the Application Locally

You'll have to go to transformers, huggingface llama3 homepage to get access to model_id = "meta-llama/Meta-Llama-3-8B" which is a gated repo to run the following code. You can also directly ask for access to following gated model: https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct . It will only take a few minutes :)

1. Start the FastAPI backend: uvicorn backend.main:app --reload

2. Start the Streamlit frontend: streamlit run frontend/main.py

Usage

• Open the Streamlit app in your browser
• Enter your preferences and generate an itinerary
• View and adjust your itinerary dynamically

Note

• Ensure Neo4j database is running and accessible.