README.md

metadata

configs:
  - config_name: default
    data_files:
      - split: train
        path:
          - '*/*.gz'
  - config_name: arxiv_abstracts
    data_files:
      - split: train
        path:
          - arxiv_abstracts/*.gz
  - config_name: arxiv_papers
    data_files:
      - split: train
        path:
          - arxiv_papers/*.gz
  - config_name: biodiversity_heritage_library
    data_files:
      - split: train
        path:
          - biodiversity_heritage_library/*.gz
  - config_name: caselaw_access_project
    data_files:
      - split: train
        path:
          - caselaw_access_project/*.gz
  - config_name: cccc
    data_files:
      - split: train
        path:
          - cccc/*.gz
  - config_name: data_provenance_initiative
    data_files:
      - split: train
        path:
          - data_provenance_initiative/*.gz
  - config_name: doab
    data_files:
      - split: train
        path:
          - doab/*.gz
  - config_name: foodista
    data_files:
      - split: train
        path:
          - foodista/*.gz
  - config_name: github_archive
    data_files:
      - split: train
        path:
          - github_archive/*.gz
  - config_name: library_of_congress
    data_files:
      - split: train
        path:
          - library_of_congress/*.gz
  - config_name: libretexts
    data_files:
      - split: train
        path:
          - libretexts/*.gz
  - config_name: news
    data_files:
      - split: train
        path:
          - news/*.gz
  - config_name: oercommons
    data_files:
      - split: train
        path:
          - oercommons/*.gz
  - config_name: peS2o
    data_files:
      - split: train
        path:
          - peS2o/*.gz
  - config_name: pre_1929_books
    data_files:
      - split: train
        path:
          - pre_1929_books/*.gz
  - config_name: pressbooks
    data_files:
      - split: train
        path:
          - pressbooks/*.gz
  - config_name: project_gutenberg
    data_files:
      - split: train
        path:
          - project_gutenberg/*.gz
  - config_name: public_domain_review
    data_files:
      - split: train
        path:
          - public_domain_review/*.gz
  - config_name: pubmed
    data_files:
      - split: train
        path:
          - pubmed/*.gz
  - config_name: python_enhancement_proposals
    data_files:
      - split: train
        path:
          - python_enhancement_proposals/*.gz
  - config_name: regulations
    data_files:
      - split: train
        path:
          - regulations/*.gz
  - config_name: stackexchange
    data_files:
      - split: train
        path:
          - stackexchange/*.gz
  - config_name: stackv2_edu
    data_files:
      - split: train
        path:
          - stackv2_edu/*.gz
  - config_name: stackv2_html
    data_files:
      - split: train
        path:
          - stackv2_html/*.gz
  - config_name: ubuntu_irc
    data_files:
      - split: train
        path:
          - ubuntu_irc/*.gz
  - config_name: uk_hansard
    data_files:
      - split: train
        path:
          - uk_hansard/*.gz
  - config_name: usgpo
    data_files:
      - split: train
        path:
          - usgpo/*.gz
  - config_name: uspto
    data_files:
      - split: train
        path:
          - uspto/*.gz
  - config_name: wikimedia
    data_files:
      - split: train
        path:
          - wikimedia/*.gz
  - config_name: wikiteam
    data_files:
      - split: train
        path:
          - wikiteam/*.gz
  - config_name: youtube
    data_files:
      - split: train
        path:
          - youtube/*.gz

Comma v0.1 dataset

This repository contains the dataset used to train Comma v0.1-1T and Comma v0.1-2T. It is a slightly modified and consolidated version of the Common Pile v0.1 "filtered" data. If you are looknig for the raw Common Pile v0.1 data, please see this collection. You can learn more about Common Pile in our paper.