Classifier with Mixed Input

Follow the full discussion on Reddit.
Hello all! I am trying to design a neural network that takes 3 inputs: a 10x10 grid with float values between 0-1, and x coordinate between 0-9, and a y coordinate between 0-9. The grid represents inputs for a number puzzles, where special data is very relevant, and the end goal would be a completed 10x10 grid of either 0 or 1 where each tile is determined by separately by the network. I don’t have much experience designing neural networks with mixed inputs, so any tips, advice, or code examples would be super helpful! There could be a much better way to structure this project, but this was what makes the most sense to me intuitively