LexSumm / README.md
CJWeiss's picture
Finally the correct README.md format (final commit)
04e614d verified
metadata
configs:
  - config_name: billsum
    data_files:
      - split: train
        path: billsum/train.jsonl
      - split: validation
        path: billsum/validation.jsonl
      - split: test
        path: billsum/test.jsonl
  - config_name: eurlexsum
    data_files:
      - split: train
        path: eurlexsum/train.jsonl
      - split: validation
        path: eurlexsum/validation.jsonl
      - split: test
        path: eurlexsum/test.jsonl
  - config_name: govreport
    data_files:
      - split: train
        path: govreport/train.jsonl
      - split: validation
        path: govreport/validation.jsonl
      - split: test
        path: govreport/test.jsonl
  - config_name: multilong
    data_files:
      - split: train
        path: multilong/train.jsonl
      - split: validation
        path: multilong/validation.jsonl
      - split: test
        path: multilong/test.jsonl
  - config_name: multishort
    data_files:
      - split: train
        path: multishort/train.jsonl
      - split: validation
        path: multishort/validation.jsonl
      - split: test
        path: multishort/test.jsonl
  - config_name: multitiny
    data_files:
      - split: train
        path: multitiny/train.jsonl
      - split: validation
        path: multitiny/validation.jsonl
      - split: test
        path: multitiny/test.jsonl
  - config_name: inabs
    data_files:
      - split: train
        path: inabs/train.jsonl
      - split: validation
        path: inabs/validation.jsonl
      - split: test
        path: inabs/test.jsonl
  - config_name: ukabs
    data_files:
      - split: train
        path: ukabs/train.jsonl
      - split: validation
        path: ukabs/validation.jsonl
      - split: test
        path: ukabs/test.jsonl

LexSumm Benchmark

The LexSumm Benchmark contains 8 datasets from various juresdictions such as the US, UK, EU, and India. Each dataset adheres to the same format with columns: input, output, and id. Below are brief descriptions of the individual datasets included in LexSumm, as well as the links to the individual datasets:

  • BillSum: US Congressional bills with summaries by the Congressional Research Service.
  • InAbs: Indian Supreme Court cases with headnotes as summaries.
  • UKAbs: UK Supreme Court judgments with official press summaries.
  • EurLexSum: Summaries of enforced EU legislation from the EUR-Lex platform.
  • GovReport: U.S. Government Accountability Office reports with expert-written summaries.
  • MultiLexSum-Long: Multi-paragraph summaries of U.S. civil rights lawsuits.
  • MultiLexSum-Short: Single-paragraph summaries of civil rights lawsuits.
  • MultiLexSum-Tiny: One-sentence summaries of civil rights lawsuits in Twitter-like format.