I am running some experiments with the Evolution Strategies algorithm at RLlib, and I need to make the results reproducible. However, setting the “seed” in the configuration file as posted here (Reproducible training - setting seeds for all workers / environments) is not working. Do you know how I can have identical runs?
The configuration file that I use is:
humanoid-v2-es:
env: Humanoid-v2
run: ES
stop:
episode_reward_mean: 200
config:
# Works for both torch and tf.
framework: tf
num_workers: 2
seed: 42
train_batch_size: 10000
Hey @Fot , ah, I think this makes sense. RLlib does all the seed handling in the RolloutWorkers, which ES and ARS don’t use (one more good reason to force-move these two into using standard Policies/RolloutWorkers/etc…).
(one more good reason to force-move these two into using standard Policies/RolloutWorkers/etc…).