Implementing DE in pytorch.optim


I’d like to use PyTorch to run Evolutionary algorithms on GPU. I noticed that DE (differential evolution) is implemented in Torch (, so I was thinking to implement it in PyTorch.

As DE is not (widely) used in Deep Learning I wanted to check with you if it would be in your interest to have DE implemented in PyTorch? I’m happy to discuss the implementation and create PR when ready.

Once I learn more about PyTorch (by implementing DE) I’d also like to implement Genetic algorithm (GA) in the near future.


You should probably post to github as such features are likely better discussed there. :slight_smile:

Ah okay, will post it there then.