Datasets:

ChenWu98
/

sibling.5.500.10.50000

dataset_info:
  features:
    - name: input_text
      dtype: string
    - name: target_text
      dtype: string
  splits:
    - name: train
      num_bytes: 3005871
      num_examples: 50000
    - name: valid
      num_bytes: 114000
      num_examples: 3000
  download_size: 2057025
  dataset_size: 3119871

Dataset Card for "sibling.5.500.10.50000"

Data examples for the paper Roll the dice & look before you leap: Going beyond the creative limits of next-token prediction