unilab.algos.mlx.ppo.ppo.PPOTrainer

class unilab.algos.mlx.ppo.ppo.PPOTrainer[source]

Bases: object

PPO update logic for MLPActorCritic and RolloutBuffer.

Parameters:

Methods

__init__(model, cfg)

update(buffer[, iteration])

__init__(model, cfg)[source]
Parameters:
update(buffer, iteration=-1)[source]
Parameters:
Return type:

Dict[str, float]