TextGAN Generator

class textbox.module.Generator.TextGANGenerator.TextGANGenerator(config, dataset)[source]

Bases: UnconditionalGenerator

The generator of TextGAN.

adversarial_loss(real_data, discriminator_func)[source]

Calculate the adversarial generator loss of real_data guided by discriminator_func.

Parameters

  • real_data (torch.Tensor) – The realistic sentence data, shape: [batch_size, max_seq_len].

  • discriminator_func (function) – The function provided from discriminator to calculated the loss of generated sentence.

Returns

The calculated adversarial loss, shape: [].

Return type

torch.Tensor

calculate_loss(corpus, nll_test=False)[source]

Calculate the generated loss of corpus.

Parameters

  • corpus (Corpus) – The corpus to be calculated.

  • nll_test (Bool) – Optional; if nll_test is True the loss is calculated in sentence level rather than in word level.

Returns

The calculated loss of corpus, shape: [].

Return type

torch.Tensor

generate(batch_data, eval_data)[source]

Generate tokens of sentences using eval_data.

Parameters

  • batch_data (Corpus) – Single batch corpus information of evaluation data.

  • eval_data – Common information of all evaluation data.

Returns

The generated tokens of each sentence.

Return type

List[List[str]]

sample()[source]

Sample a batch of generated sentence indice.

Returns

The generated sentence indice, shape: [batch_size, max_length]. torch.Tensor: The latent code of the generated sentence, shape: [batch_size, hidden_size].

Return type

torch.Tensor

training: bool