In the library PyMetaheuristic there is a function called whale_optimization_algorithm. From my knowledge it is a simple yet powerful library which doesn't result in dependency hell. If I want to use that for feature selection of a large dataset, it would be heavy on cpu and would require a lot of time to complete.
Can this function be made GPU dependent so that time can be reduced ?
Also if this is to be done in Kaggle or Colab how to do that as well ?