We've just built a tutorial for TensorFlow on CodinGame.
Our platform currently offers free multiplayer games that can be solved using various AI methods but we would like to enable our users to go further into machine learning. So we've installed TensorFlow and integrated the MNIST tutorial to our platform (in particular, we've added a visualization module that shows your wrong guesses).
Our goal with this tutorial is 1) diversify the platform uses towards ML 2) allow developers interested to learn about TF to benefit from a user-friendly infrastructure to do so.
You can have a look at the prototype here: https://goo.gl/POAIjp
Do you see any practical uses for this tutorial? Or any paths for improvement?
Any feedback would help us a lot. Thanks for checking!