Text Summarization using sequence to sequence based model with attention mechanism and beam search inference
The dataset used is a subset of the gigaword dataset and can be found at here It contains 3,803,955 parallel source & target examples for training and 189,649 examples for validation out of which 80000 is used for training and 1000 for validation.