metadata

languages:
  - en
multilinguality:
  - monolingual
size_categories:
  - 10K<n<100K
task_categories:
  - conditional-text-generation
task_ids:
  - summarization

GovReport dataset for summarization

Dataset for summarization of long documents.
Adapted from this repo.
Note that original data are pre-tokenized so this dataset returns " ".join(text).
This dataset is compatible with the run_summarization.py script from Transformers if you add this line to the summarization_name_mapping variable:

"ccdv/govreport-summarization": ("report", "summary")

Data Fields

id: paper id
report: a string containing the body of the report
summary: a string containing the summary of the report

Data Splits

This dataset has 3 splits: train, validation, and test.
Token counts with a RoBERTa tokenizer.

Dataset Split	Number of Instances	Avg. tokens
Train	17,517	< 9,000 / < 500
Validation	973	< 9,000 / < 500
Test	973	< 9,000 / < 500

Cite original article

@misc{huang2021efficient,
      title={Efficient Attentions for Long Document Summarization}, 
      author={Luyang Huang and Shuyang Cao and Nikolaus Parulian and Heng Ji and Lu Wang},
      year={2021},
      eprint={2104.02112},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
    }