espnet2.uasr.loss.gradient_penalty.UASRGradientPenalty

source

class espnet2.uasr.loss.gradient_penalty.UASRGradientPenalty(discriminator: AbsDiscriminator, weight: float = 1.0, probabilistic_grad_penalty_slicing: str2bool = False, reduction: str = 'sum')

Bases: AbsUASRLoss

gradient penalty for UASR.

Initializes internal Module state, shared by both nn.Module and ScriptModule.

forward(fake_sample: Tensor, real_sample: Tensor, is_training: str2bool, is_discrimininative_step: str2bool)

Forward.

• Parameters:
  • fake_sample – generated sample from generator
  • real_sample – real sample
  • is_training – whether is at training step
  • is_discriminative_step – whether is training discriminator