Update README.md

README.md

@@ -1,99 +1,11 @@
-# Prompt Search Engine
-
-## Overview
-
-The Prompt Search Engine is a Flask-based web application designed to search for the most similar prompts from a dataset using cosine similarity. It leverages Hugging Face's `sentence-transformers` to vectorize prompts and stores them in a Pinecone vector database for efficient querying. The frontend is built using Streamlit, providing an intuitive interface for users to input their queries and get results.
-
-## Features
-
-- **Efficient Vector Search**: Uses Pinecone for storing and querying vector embeddings.
-- **Cosine Similarity Calculation**: Custom implementation for calculating cosine similarity between query vectors and stored vectors.
-- **Streamlit Interface**: Simple and user-friendly interface for querying and displaying results.
-- **Logging**: Comprehensive logging for easy debugging and monitoring.
-
-## Prerequisites
-
-- Python 3.9 or higher
-- Pinecone API key
-
-## Setup
-
-### Clone the Repository
-
-```
-git clone https://github.com/your-username/prompt-search-engine.git
-cd prompt-search-engine
-```
-## Create a Virtual Environment and Install Dependencies
-```
-python3 -m venv venv
-source venv/bin/activate
-pip install -r requirements.txt
-```
-
-## Configure Pinecone
-```
-# Replace YOUR_PINECONE_API_KEY with your actual Pinecone API key in the vectorizer.py file.
-python
-pinecone = Pinecone(api_key='YOUR_PINECONE_API_KEY')
-```
-### Initial Data Load
-```
-# Run the script to load the initial dataset into the Pinecone database:
-
-python load_data.py
-```
-
-### Run the Flask Backend
-```
-python run.py
-```
-
-### Run the Streamlit Frontend
-```
-# In a new terminal (while the backend is running), start the Streamlit app:
-
-streamlit run ui/app.py
-```
-### Running the Tests 
-To run the tests, navigate to the root directory and execute:
-```
-python run_tests.py
-```
-Make sure that you executed python load_data.py before.
-You should receive something like this:
-<img width="576" alt="image" src="https://github.com/user-attachments/assets/a9cd8acb-9280-4b55-9bec-3009a0a61b87">
-
-### Rebuild and run Docker container
-```
-docker build -t prompt-search-engine .
-docker run -p 5000:5000 prompt-search-engine
-```
-
-### Usage
-Open your web browser and go to http://localhost:8501.
-Enter a query in the input box.
-Adjust the number of results using the slider.
-Click "Search" to get the most similar prompts from the dataset.
-
-### File Descriptions
-##### app/: Contains the core logic for vectorization and search functionality.
-##### search_engine.py: Implements the PromptSearchEngine class for querying the Pinecone database.
-##### vectorizer.py: Implements the Vectorizer class for loading and storing vectors in Pinecone.
-##### ui/: Contains the Streamlit frontend application.
-##### app.py: Streamlit app for user interface.
-##### load_data.py: Script to load the initial dataset into Pinecone.
-##### run.py: Flask application entry point.
-##### requirements.txt: Lists the Python dependencies for the project.
-
-### Logging
-Logging is configured in the vectorizer.py and search_engine.py files.
-
-### License
-This project is licensed under the MIT License.
-
-## Acknowledgements
-##### - Hugging Face for sentence-transformers
-##### - Pinecone for vector database services
-##### - Streamlit for the web app interface
-##### - Feel free to reach out if you have any questions or need further assistance. Enjoy using the Prompt Search Engine!
+---
+title: Prompt Search Engine
+emoji: 🌍
+colorFrom: yellow
+colorTo: blue
+sdk: docker
+sdk_version: 4.38.1
+app_file: app.py
+pinned: false
+license: mit
+---