NLP_practice

vanilla_rnn

Three different kinds of implement of rnn, including hand-made, static_rnn and dynamic_rnn.

Train a seq2seq autoencoder with self-generated toy data.

Self-implemented code used in 20180409 MSLAB group meeting.

Name		Name	Last commit message	Last commit date
Latest commit History 17 Commits
README.md		README.md
autoencoder_lstm_embedding.ipynb		autoencoder_lstm_embedding.ipynb
autoencoder_lstm_one_hot.ipynb		autoencoder_lstm_one_hot.ipynb
dqn.ipynb		dqn.ipynb
imdb_ae_lstm.ipynb		imdb_ae_lstm.ipynb
imdb_ae_lstm_simultaneous.ipynb		imdb_ae_lstm_simultaneous.ipynb
imdb_ae_lstm_simultaneous_double_vocab.ipynb		imdb_ae_lstm_simultaneous_double_vocab.ipynb
imdb_bag_of_word.ipynb		imdb_bag_of_word.ipynb
nltk_brown_ae_bilstm.ipynb		nltk_brown_ae_bilstm.ipynb
nltk_brown_ae_lstm.ipynb		nltk_brown_ae_lstm.ipynb
requirements.txt		requirements.txt
vanilla_rnn.ipynb		vanilla_rnn.ipynb