I‘m looking for a Python implementation of a Restricted Boltzmann Machine (RBM), e.g. applied to MNIST data as mentioned in „Elements of Statistical Learning“ Ch. 17, in Tensorflow 2.x.
I‘m aware of code as linked here. However, the model(s) are implemented in TF 1 and some layers are not supported any more (in TF2). Some links are also broken.
Can someone point me to a TF2 implementation of RBM or alternatively some implementation using sklearn
?