TanhDelta

class torchrl.modules.TanhDelta(param: torch.Tensor, low: torch.Tensor | float = - 1.0, high: torch.Tensor | float = 1.0, event_dims: int = 1, atol: float = 1e-06, rtol: float = 1e-06, safe: bool = True)

Implements a Tanh transformed_in Delta distribution.

Parameters:

• param (torch.Tensor) – parameter of the delta distribution;

• low (torch.Tensor or number, optional) – minimum value of the distribution. Default is -1.0;

• high (torch.Tensor or number, optional) – maximum value of the distribution. Default is 1.0;

• event_dims (int, optional) – number of dimensions describing the action. Default is 1;

• atol (number, optional) – absolute tolerance to consider that a tensor matches the distribution parameter; Default is 1e-6

• rtol (number, optional) – relative tolerance to consider that a tensor matches the distribution parameter; Default is 1e-6

• batch_shape (torch.Size, optional) – batch shape;

• event_shape (torch.Size, optional) – shape of the outcome;

property mean: Tensor

Returns the mean of the distribution.

property mode: Tensor

Returns the mode of the distribution.