Home

Awesome

Sequicity

Source code for the ACL 2018 paper entitled "Sequicity: Simplifying Task-oriented Dialogue Systems with Single Sequence-to-Sequence Architectures" by Wenqiang Lei et al.

@inproceedings{lei2018sequicity,
  title={Sequicity: Simplifying task-oriented dialogue systems with single sequence-to-sequence architectures},
  author={Lei, Wenqiang and Jin, Xisen and Kan, Min-Yen and Ren, Zhaochun and He, Xiangnan and Yin, Dawei},
  booktitle={ACL},
  year={2018}
}

Training with default parameters

python model.py -mode train -model [tsdf-camrest|tsdf-kvret]

(optional: configuring hyperparameters with cmdline)

python model.py -mode train -model [tsdf-camrest|tsdf-kvret] -cfg lr=0.003 batch_size=32

Testing

python model.py -mode test -model [tsdf-camrest|tsdf-kvret]

Reinforcement fine-tuning

python model.py -mode rl -model [tsdf-camrest|tsdf-kvret] -cfg lr=0.0001

Before running

  1. Install required python packages. We used pytorch 0.3.0 and python 3.6 under Linux operating system.
pip install -r requirements.txt
  1. Make directories under PROJECT_ROOT.
mkdir vocab
mkdir log
mkdir results
mkdir models
mkdir sheets
  1. Download pretrained Glove word vectors and place them in PROJECT_ROOT/data/glove.