Datasets:

nanotron
/

simple_needle_in_a_hay_stack

dataset_info:
  features:
    - name: id
      dtype: string
    - name: prompt
      dtype: string
    - name: answer
      dtype: int64
    - name: context_length
      dtype: int64
    - name: depth_percent
      dtype: float64
  splits:
    - name: train
      num_bytes: 139407372
      num_examples: 12600
  download_size: 8384165
  dataset_size: 139407372
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*