Awesome

entailment_bank

This repository contains evaluation code for the paper Explaining Answers with Entailment Trees (EMNLP 2021), see below.

See entailer.md for information on how to run the Entailer model from follow-up paper: Entailer: Answering Questions with Faithful and Truthful Chains of Reasoning (EMNLP 2022).

EntailmentBank evaluation code

Dataset available at https://allenai.org/data/entailmentbank

This dataset is also published in a more readable format as part of this book: http://cognitiveai.org/dist/entailmentbank-book-may2022.pdf

EntailmentBank annotation tool can be found at https://github.com/cognitiveailab/entailmentbank-tree-annotation-tool

Setting up python environment

• conda create -n entbank python=3.7
• conda activate entbank
• pip install -r requirements.txt
• Download the bleurt-large-512 model from https://github.com/google-research/bleurt/blob/master/checkpoints.md

Example Evaluation Commands:

Task1:

python eval/run_scorer.py \
  --task "task_1" \
  --split test \
  --prediction_file PREDICTION-TSV-PATH  \
  --output_dir  OUTPUT-PATH  \
  --bleurt_checkpoint "PATH to bleurt-large-512 model"

Task2:

python eval/run_scorer.py \
  --task "task_2" \
  --split test \
  --prediction_file PREDICTION-TSV-PATH  \
  --output_dir  OUTPUT-PATH  \
  --bleurt_checkpoint "PATH to bleurt-large-512 model" 

Task3:

python eval/run_scorer_task3.py \
  --split test \
  --prediction_file PREDICTION-TSV-PATH  \
  --output_dir  OUTPUT-PATH  \
  --bleurt_checkpoint "PATH to bleurt-large-512 model" 

Prediction file format

Prediction file (PREDICTION-TSV-PATH) is a single column TSV file with datapoints in the same order as the public dataset jsonl file. Value of each line is the predicted proof in the DOT format. e.g.

$proof$ = sent2 & sent3 -> int1: the northern hemisphere is a kind of place; int1 & sent1 -> hypothesis;

For task 3, when using a custom retrieved set of sentences, the prediction file should instead be a jsonl file where each line includes a 'worldtree_provenance' field specifying what each sentence refers to, as well as an 'id' and 'slots' + 'proof' field, following the format in this example:

{"id":"Mercury_SC_401371","slots":{"proof":"sent23 & sent3 -> int1: the sun rising and setting are kinds of events; int1 & sent4 -> hypothesis;"},
 "worldtree_provenance":{"sent12":{"uuid":"a972-0cbb-3c14-b098","original_text":"cycles of day and night occur once per day"},
   "sent4":{"uuid":"f8be-9407-d60e-6282","original_text":"the sun rising; setting occurs once per day"},...}}

Citation

@article{entailmentbank2021,
  title={Explaining Answers with Entailment Trees},
  author={Dalvi, Bhavana and Jansen, Peter and Tafjord, Oyvind and Xie, Zhengnan and Smith, Hannah and Pipatanangkura, Leighanna and Clark, Peter},
  journal={EMNLP},
  year={2021}
}