Hello I am tryning to solve a simple optimization for a million times with different inputs. For that I have the following structure:
input = list of N arrays myfun() = Takes one array solves and returns another list as answer
Now I want to parallelize this over three nodes, where each node has 48 CPUs.
Can anyone tell me how to proceed?
Currently, I can run a joblib parallel delayed function, but that can only work over a single node.
Currently, I can run a joblib parallel delayed function, but that can only work over a single node.