Pile_Tokenized / README.md
Trevor Dohm
Upload dataset (part 00047-of-00048)
89434b6 verified
metadata
dataset_info:
  features:
    - name: text
      dtype: string
    - name: input_ids
      sequence: int32
    - name: attention_mask
      sequence: int8
  splits:
    - name: train
      num_bytes: 1197349778214
      num_examples: 134318121
  download_size: 606495022503
  dataset_size: 1197349778214
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*