Tic-Tac-Toe

Tic-Tac-Toe Probable winner

My first attempt to create model and to learn TensorFlow. Goal was to create NN, which take tic-tac-toe board and output the winner.

Web application: Huggingface Space - Tic-Tac-Toe Probable Winner
Jupyter notebook for trainig: Datalore - Tic-Tac-Toe Neural Network

Neural network

Input is a tensor of shape (9, 3)

  • 9 represents the tic-tac-toe board
  • 3 represents the state of the box:
    • [1, 0, 0] for O Player
    • [0, 1, 0] for X Player
    • [0, 0, 1] for empty space

Output is a tensor of shape (3):

  • Index 0 represents 0 player win
  • Index 1 represents X player win
  • Index 2 represents that there is no winning condition
Tic-Tac-Toe
Downloads last month
-
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support