fubuki119's picture
Tokenized dataset
d7e175b
metadata
dataset_info:
  features:
    - name: input_ids
      sequence: int32
    - name: attention_mask
      sequence: int8
    - name: labels
      sequence: int64
  splits:
    - name: train
      num_bytes: 429594853
      num_examples: 1659083
    - name: validation
      num_bytes: 174392
      num_examples: 520
    - name: test
      num_bytes: 992781
      num_examples: 2507
  download_size: 139343444
  dataset_size: 430762026
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*
      - split: test
        path: data/test-*