Shortcuts

Policy Gradient Methods

Loss modules for policy gradient algorithms.

PPOLoss(*args, **kwargs)

A parent PPO loss class.

ClipPPOLoss(*args, **kwargs)

Clipped PPO loss.

KLPENPPOLoss(*args, **kwargs)

KL Penalty PPO loss.

A2CLoss(*args, **kwargs)

TorchRL implementation of the A2C loss.

ReinforceLoss(*args, **kwargs)

Reinforce loss module.

Docs

Lorem ipsum dolor sit amet, consectetur

View Docs

Tutorials

Lorem ipsum dolor sit amet, consectetur

View Tutorials

Resources

Lorem ipsum dolor sit amet, consectetur

View Resources