Snippets - gugarosa/nalp GitHub Wiki

Our code belongs to everyone. Thus, we strive to offer the most possible commented, documented, and exemplified code of all time. In NALP, we present code snippets in an attempt to fulfill everyone's needs.

Applications

• adversarial_char_generation.py: Generate characters with text-based adversarial models;
• adversarial_image_generation.py: Generate images with adversarial models;
• adversarial_word_generation.py: Generate words with text-based adversarial models;
• recurrent_audio_generation.py: Generate audio with recurrent models;
• recurrent_char_generation.py: Generate characters with recurrent models;
• recurrent_word_generation.py: Generate words with recurrent models.

Corpus

• create_audio_corpus.py: AudioCorpus class creation;
• create_sentence_corpus.py: SentenceCorpus class creation;
• create_text_corpus.py: TextCorpus class creation.

Datasets

• create_image_dataset.py: ImageDataset class creation;
• create_language_modelling_dataset.py: LanguageModellingDataset class creation.

Encoders

• encode_with_integer.py: How-to encode tokens with IntegerEncoder;
• encode_with_word2vec.py: How-to encode tokens with Word2vecEncoder.

Models

Adversarial

• train_dcgan.py: DCGAN training;
• train_gan.py: GAN training;
• train_gsgan.py: GSGAN training;
• train_maligan.py: MaliGAN training;
• train_relgan.py: RelGAN training;
• train_seqgan.py: SeqGAN training;
• train_wgan.py: WGAN training;
• train_wgan_gp.py: WGAN-GP training.

Generators

• train_bi_lstm.py: Bidirectional LSTM training;
• train_gru.py: GRU training;
• train_lstm.py: LSTM training;
• train_rmc.py: RMC training;
• train_rnn.py: RNN training;
• train_stacked_rnn.py: Stacked RNN training.

Utils

• load_audio_data.py: How-to load audio data;
• load_sentence_data.py: How-to load sentence-based data;
• load_text_data.py: How-to load text data;
• preprocess_sentence_data.py: Pre-processings steps when using sentenced-based data;
• preprocess_text_data.py: Pre-processings steps when using text data.