Forward-Forward is an algorithm that was unveiled in NeurIPS '22. Hinton's Lecture [https://www.cs.toronto.edu/~hinton/FFA13.pdf] I saw a lot of implementations of this algorithm in pytorch, I personally like tensorflow so I tried to recreate the algorithm using pytorch. I was inspired by [https://github.com/madcato/forward-forward-pytorch] repo. And on top of that, I tried to recreate it in Tensorflow.
Status: This is an on-going project, I aim to deliver more updates and implementations of this algorithm beyond the scope of Conv networks. (Though I used a dense layer for the current implementation) and with other datasets.
Contribution: Feel free to contribute to the project so that we can make forward-forward the default algorithm used in Tensorflow :)