metadata
dataset_info:
features:
- name: transcript
dtype: string
- name: num_moves
dtype: int8
- name: checkmate
dtype: bool
splits:
- name: train
num_bytes: 5082951710.194961
num_examples: 14399003
- name: val
num_bytes: 564772686.8050389
num_examples: 1599890
download_size: 3358564045
dataset_size: 5647724397
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: val
path: data/val-*