---
dataset_info:
  features:
  - name: input
    sequence: int64
  - name: output
    sequence: int64
  - name: turn
    dtype: string
  - name: winner
    dtype: string
  splits:
  - name: train
    num_bytes: 358159550
    num_examples: 646560
  download_size: 8580634
  dataset_size: 358159550
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---