Reinforcement Learning Algorithms with Python

Parallelizing ES

It's easy to see how ES can be scaled across multiple CPUs: each worker is assigned to a separate candidate solution of the population. The evaluation can be done in complete autonomy, and as described in the paper, optimization can be done in parallel on each worker, with only a few scalars shared between each CPU unit.

Specifically, the only information that's shared between workers is the scalar return, , of an episode and the random seed that has been used to sample . The amount of data can be further shrunk by sending only ...