Implentation of 'Variational Attention for Sequence to Sequence Models' in tensorflow.

This package consists of 3 models, each of which have been organized into separate folders:

1. Deterministic encoder-decoder with deterministic attention (ded_detAttn)
2. Variational encoder-decoder with deterministic attention (ved_detAttn)
3. Variational encoder-decoder with variational attention (ved_varAttn)

The proposed model and baselines have been evaluated on two experiments:

1. Neural Question Generation with the SQuAD dataset
2. Conversation Systems with the Cornell Movie Dialogue dataset

The data has been preprocessed and the train-val-test split is provided in the data/ directory.

• tensorflow-gpu==1.3.0
• Keras==2.0. 8
• numpy==1.12.1
• pandas==0.22.0
• gensim==3.1.2
• nltk==3.2.3
• tqdm==4.19.1

1. Generate word2vec, required for initializing word embeddings, specifying the dataset:

python w2v_generator.py --dataset qgen 

2. Train the desired model, set configurations in the model_config.py file. For example,

cd ved_varAttn
vim model_config.py # Make necessary edits
python train.py

• The model checkpoints are stored in models/ directory, the summaries for Tensorboard are stored in summary_logs/ directory. As training progresses, the metrics on the validation set are dumped intolog.txt and bleu/ directory.

3. Evaluate performance of the trained model. Refer to predict.ipynb to load desired checkpoint, calculate performance metrics (BLEU and diversity score) on the test set, and generate sample outputs.