Skip to content

Anonymized code for conference submission of the following paper - "A Computational Interface to Translate Unstructured Commander's Intent into a Machine Readable Specification"

anonymousturtle433/Anonymized-Code

Repository files navigation

NL_Constraints

Code for learning goals and constraints from language project

Start with installing the setup file to ensure code used in this repository is imported correctly

pip install -e .

If installing directly gives you import issues, please run the following command instead

python setup.py develop

Install Requirements

conda create --name nl_constraints python=3.8.5
conda activate nl_constraints
pip install -r requirements.txt

###Data

Keep all the json files containing training data in the folder Risk_NL_Commanders_Intent/NL_constraints_data_prep/Output_Data

The file names should be kept as -

nl_goals_constraints_{train/val/test}_{suffix}.json

The suffix depends on the type of data being used. It should be -

human = Human Data
syn = Synthetic Data 
aug = Augemeted Data
human_aug = Human Augmented Data 
synthetic_sug = Synthetic Augmented Data

About

Anonymized code for conference submission of the following paper - "A Computational Interface to Translate Unstructured Commander's Intent into a Machine Readable Specification"

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published