Self-Adaptive Control of Temperature

This is the code for our paper Learning When to Concentrate or Divert Attention: Self-Adaptive Attention Temperature for Neural Machine Translation, https://www.aclweb.org/anthology/papers/D/D18/D18-1331/

---

Requirements

• Ubuntu 16.0.4
• Python 3.6
• Pytorch >= 0.4
• pyrouge

---

Preprocessing

python3 preprocess.py -load_data path_to_data -save_data path_to_store_data 

Remember to put the data into a folder and name them train.src, train.tgt, valid.src, valid.tgt, test.src and test.tgt, and make a new folder inside called data

---

Training

python3 train.py -log log_name -config config_yaml -gpus id

---

Evaluation

python3 train.py -log log_name -config config_yaml -gpus id -restore checkpoint -mode eval

---

Citation

If you use this code for your research, please cite the paper this code is based on: Learning When to Concentrate or Divert Attention: Self-Adaptive Attention Temperature for Neural Machine Translation:.

@inproceedings{SACT,
    title = "Learning When to Concentrate or Divert Attention: Self-Adaptive Attention Temperature for Neural Machine Translation",
    author = "Lin, Junyang  and
      Sun, Xu  and
      Ren, Xuancheng  and
      Li, Muyu  and
      Su, Qi",
    booktitle = "Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing",
    month = oct # "-" # nov,
    year = "2018",
    address = "Brussels, Belgium",
    publisher = "Association for Computational Linguistics",
    url = "https://www.aclweb.org/anthology/D18-1331",
    pages = "2985--2990"
}