MaliGAN Discriminator¶
- class textbox.module.Discriminator.MaliGANDiscriminator.MaliGANDiscriminator(config, dataset)[source]¶
Bases:
UnconditionalGenerator
MaliGANDiscriminator is LSTMs.
- calculate_loss(real_data, fake_data)[source]¶
Calculate the loss for real data and fake data. The discriminator is trained with the standard objective that GAN employs.
- Parameters
real_data (torch.Tensor) – The realistic sentence data, shape: [batch_size, max_seq_len].
fake_data (torch.Tensor) – The generated sentence data, shape: [batch_size, max_seq_len].
- Returns
The calculated loss of real data and fake data, shape: [].
- Return type
torch.Tensor
- forward(data)[source]¶
Calculate the probability that the data is realistic.
- Parameters
data (torch.Tensor) – The sentence data, shape: [batch_size, max_seq_len].
- Returns
The probability that each sentence is realistic, shape: [batch_size].
- Return type
torch.Tensor
- training: bool¶