LemonadeAPI: Time Series Forecasting

Welcome to LemonadeAPI, an advanced time series forecasting service leveraging PyTorch-Forecasting and Temporal Fusion Transformers. This project is currently in the early stages and focuses on forecasting stock prices.

Features

• Model: Uses state-of-the-art Temporal Fusion Transformers from the PyTorch-Forecasting package.
• Ease of Use: Simple setup via requirements.txt.
• Scalability: Designed with scalability in mind, we aim to extend this service to handle multiple stocks and indices.
• Deployment: Pre-configured for deployment on AWS Lambda for serverless computing and effortless scaling.

Installation

To get started with LemonadeAPI, you need to clone this repository and install the requirements. Assuming you have Python installed, you can do this by following these steps:

    git clone https://github.com/yourusername/LemonadeAPI.git
    cd LemonadeAPI
    pip install -r requirements.txt

Usage

Currently, the project includes a Jupyter notebook that trains a Temporal Fusion Transformer model to predict AMD stock prices.

Future Directions

This project is in its infancy, and we have big plans for its future. We aim to incorporate automated training processes for given data and to broaden the scope of forecasts to include a wide array of stocks.

Contributing

We welcome contributions to LemonadeAPI! Please feel free to submit pull requests, or open an issue to discuss potential changes or features you'd like to see.

Support

If you encounter any bugs or issues with the current implementation, please feel free to raise an issue on GitHub.