DistributedReinforcementLearning.TrainerType
Trainer(;policy, sealer=deepcopy)

A wrapper around AbstractPolicy, the sealer is used to create an immutable object from the inner parameters of policy when received a FetchParamMsg.

source