NL2Bash contains 12k one-line Linux shell commands used in practice and their natural language descriptions provided by experts. The dataset includes 100+ commonly used shell utilities, providing a rich training and test bed for automatically translating natural language into command lines.

The parallel data are stored with a source file bash/\*.nl and a target file bash/\*.cm. The entire corpus bash/all.nl, bash/all.cm are randomly splited into train, dev, and test set with a ratio of 10:1:1.

The manual evaluations of a subset of model predictions are stored in bash/manual_judgments/ in csv format.

Filtering the raw parallel corpus.

python3 scripts/filter_data.py bash

Randomly split the filtered parallel corpus into train, dev and test (using fixed random seed).

python3 scripts/split_data.py bash

Statistics of the dataset split is shown below.

cd ../experiments

./bash-run.sh --process_data --dataset bash

If you use the data or source code in your work, please cite

@inproceedings{LinWZE2018:NL2Bash, 
  author = {Xi Victoria Lin and Chenglong Wang and Luke Zettlemoyer and Michael D. Ernst}, 
  title = {NL2Bash: A Corpus and Semantic Parser for Natural Language Interface to the Linux Operating System}, 
  booktitle = {Proceedings of the Eleventh International Conference on Language Resources
               and Evaluation {LREC} 2018, Miyazaki (Japan), 7-12 May, 2018.},
  year = {2018} 
}