README.md

metadata

dataset_info:
  features:
    - name: tokens
      sequence: uint16
    - name: is_memorized
      dtype: bool
    - name: index
      dtype: int64
  splits:
    - name: train
      num_bytes: 20540625000
      num_examples: 5000000
  download_size: 17785569040
  dataset_size: 20540625000

Dataset Card for "pile-duped-pythia-random-sampled"

More Information needed