virne.solver.learning.rl_base.shared_adam

Functions

sync_gradients(shared_model, local_model)

Classes

SharedAdam(params[, lr, betas, eps, ...])

Shared optimizer, the parameters in the optimizer will shared in the multiprocessors.