`torch_mist.critic.base`

Module Contents

Classes

Attributes

`CRITIC_TYPE`
`SEPARABLE_CRITIC`
`JOINT_CRITIC`
`CRITIC_TYPES`

torch_mist.critic.base.CRITIC_TYPE = 'critic_type'

torch_mist.critic.base.SEPARABLE_CRITIC = 'separable'

torch_mist.critic.base.JOINT_CRITIC = 'joint'

torch_mist.critic.base.CRITIC_TYPES

class torch_mist.critic.base.Critic

Bases: torch.nn.Module

abstract forward(x: torch.Tensor, y: torch.Tensor) → torch.Tensor: Compute the value of the critic evaluated at the pair (x, y) :param x: a tensor representing x :param y: a tensor representing y :return: The value of the ratio estimator on the given pair