Open mohitsharma29 opened 2 years ago
We don't have an implementation in PyTorch. The current implementation uses Theano to effectively optimize the mathematical operations involved in max-entropy distribution computation and should be feasible with Tensorflow frameworks if that would be useful for your applications.
Hey
Do you have the implementation of this technique in PyTorch or Numpy? It will be easy for us to use this library with other things if we can port this to PyTorch.