mtg-coloridentity / README.md
Joshua Sundance Bailey
initial commit
171c1a6
|
raw
history blame
3.17 kB
metadata
title: mtg-coloridentity
emoji: 🧙
colorFrom: white
colorTo: red
sdk: streamlit
sdk_version: 1.30.0
app_file: app.py
pinned: true
license: mit

mtg-coloridentity

License: MIT python

Push to HuggingFace Space Open HuggingFace Space

pre-commit Ruff Checked with mypy Code style: black

security: bandit

mtg-coloridentity

🤖 This README was written by GPT-4. 🤖

Overview

This Streamlit app is designed for the multi-label classification of Magic: The Gathering (MTG) cards, specifically focusing on their color identity. It utilizes a pre-trained model hosted on Hugging Face, joshuasundance/mtg-coloridentity-multilabel-classification, to predict the color identity of MTG cards based on their names and descriptions.

Features

  • Interactive UI: Users can input the name and text of any MTG card to get predictions on its color identity.
  • Color Probabilities: The app displays the probability of each color identity (Black, Green, Red, Blue, White) for the given card.
  • Random Card Selection: With a "Roll the Dice" feature, users can load the text of a random MTG card from the dataset.

How It Works

The app fetches a pre-trained SetFit model from Hugging Face and uses it to predict the color identities of MTG cards. The model's predictions are displayed as a bar chart, showing the probability of each color identity.

Getting Started

To run this app locally, clone the repository and ensure you have the following prerequisites installed:

  • Python 3.x
  • streamlit
  • pandas
  • seaborn
  • matplotlib
  • datasets and setfit from Hugging Face

Contributions, Support, and Contact

Contributions to this project are welcome! Please feel free to submit issues and pull requests.

For support, please raise an issue on GitHub or in the HuggingFace space.

License

This project is under the MIT License.

Acknowledgments

Thanks to HuggingFace and setfit!

TODO

  • make a todo list ;)
  • improve READMEs
  • make better model(s)
  • learn in public