README.md

metadata

dataset_info:
  features:
    - name: question
      dtype: string
    - name: answer
      dtype: string
    - name: instruction
      dtype: string
    - name: qtype
      dtype: string
    - name: pubid
      dtype: int32
  splits:
    - name: train
      num_bytes: 9543193566
      num_examples: 21608801
  download_size: 5032171833
  dataset_size: 9543193566
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*