yuanhangsu / elstm-dbrnn

official tensorflow implementation of "On Extended Long Short-term Memory and Dependent Bidirectional Recurrent Neural Network"

On Extended Long Short-term Memory and Dependent Bidirectional Recurrent Neural Network

This is the official implementation of extended LSTM (ELSTM) and dependent BRNN (BRNN) in paper: "On Extended Long Short-term Memory and Dependent Bidirectional Recurrent Neural Network"

Dependencies

• Python 3.6
• Tensorflow==1.15.0

Dataset

• PTB Dataset

It is the PTB dataset for language modeling

Download the dataset, create a folder named data and unzip the files to it

HOW TO

#Training

   python    generate_sequence.py --mode=train    

#Testing

   python    generate_sequence.py --mode=inference    

ELSTM

DBRNN

Citations

If you find this work useful, please consider citing it.

@article{ELSTM,
  title={On Extended Long Short-term Memory and Dependent Bidirectional Recurrent Neural Network},
  author={Su, Yuanhang and Kuo, C.-C. Jay},
  journal={Neurocomputing},
  volume={356},
  month={Sep},
  year={2019},
  pages={151-161}
}