Datasets:

andersonbcdefg
/

reward-modeling-short-tokenized

dataset_info:
  features:
    - name: preferred_input_ids
      sequence: int64
    - name: preferred_attention_masks
      sequence: int64
    - name: dispreferred_input_ids
      sequence: int64
    - name: dispreferred_attention_masks
      sequence: int64
  splits:
    - name: train
      num_bytes: 8509513392
      num_examples: 259563
  download_size: 138519630
  dataset_size: 8509513392

Dataset Card for "reward-modeling-short-tokenized"

More Information needed