Skip to content

Latest commit

 

History

History
10 lines (7 loc) · 440 Bytes

File metadata and controls

10 lines (7 loc) · 440 Bytes

seq2seq_chatbot_encoder_decoder_model_with_attention

Sequence to Sequence Encoder Decoder Model with Attention and Beam Search decoding (implemented Greedy Search Decoder as well for comparison) Implemented for chatbot but can be used for Text Summarization as well Used Cornell Movie Subtitles Dataset for training

Based on the following paper Generating News Headlines with Recurrent Neural Networks

https://arxiv.org/abs/1512.01712