Datasets:

Jedalc
/

imdb_mlm_downsampled

dataset_info:
  features:
    - name: input_ids
      sequence: int32
    - name: attention_mask
      sequence: int8
    - name: word_ids
      sequence: int64
    - name: labels
      sequence: int64
  splits:
    - name: train
      num_bytes: 27200000
      num_examples: 10000
    - name: test
      num_bytes: 2720000
      num_examples: 1000
  download_size: 6741894
  dataset_size: 29920000

Dataset Card for "imdb_mlm_downsampled"

More Information needed