reference/generated/torchrl.trainers.algorithms.configs.modules.TanhNormalModelConfig

Run in Google Colab

Colab

Download Notebook

Notebook

View on GitHub

GitHub

torchrl.trainers.algorithms.configs.modules.TanhNormalModelConfig

class torchrl.trainers.algorithms.configs.modules.TanhNormalModelConfig(_partial_: bool = False, in_keys: Any = None, out_keys: Any = None, shared: bool = False, network: MLPConfig = '???', eval_mode: bool = False, extract_normal_params: bool = True, scale_mapping: str = 'biased_softplus_1.0', scale_lb: float = 0.0001, low: Any = None, high: Any = None, tanh_loc: bool = False, param_keys: Any = None, exploration_type: Any = 'RANDOM', return_log_prob: bool = False, _target_: str = 'torchrl.trainers.algorithms.configs.modules._make_tanh_normal_model')

A class to configure a TanhNormal model.

Example

>>> cfg = TanhNormalModelConfig(network=MLPConfig(in_features=10, out_features=5, depth=2, num_cells=32))
>>> net = instantiate(cfg)
>>> y = net(torch.randn(1, 10))
>>> assert y.shape == (1, 5)

Parameters:

• low – lower bound of the action support handed to TanhNormal (a scalar or a per-dimension sequence). Defaults to None, i.e. the distribution default of -1.

• high – upper bound of the action support. Defaults to None, i.e. 1.

• tanh_loc – if True, the location is squashed to [-upscale, upscale] before the tanh transform, which keeps the log-probability of actions at the bounds finite (see TanhNormal). Defaults to False.

See also

torchrl.modules.TanhNormal