reference/generated/torchrl.trainers.algorithms.configs.utils.AdamaxConfig

Run in Google Colab

Colab

Download Notebook

Notebook

View on GitHub

GitHub

torchrl.trainers.algorithms.configs.utils.AdamaxConfig

class torchrl.trainers.algorithms.configs.utils.AdamaxConfig(lr: float = 0.002, betas: tuple[float, float] = (0.9, 0.999), eps: float = 1e-08, weight_decay: float = 0.0, foreach: bool | None = None, maximize: bool = False, differentiable: bool = False, capturable: bool = False, _target_: str = 'torch.optim.Adamax', _partial_: bool = True)

Hydra configuration for torch.optim.Adamax.

Every kwarg accepted by torch.optim.Adamax.__init__ is exposed as a field here.