metadata
dataset_info:
features:
- name: id
dtype: string
- name: template_args
struct:
- name: character
dtype: string
- name: op1
dtype: int64
- name: op2
dtype: int64
- name: result
dtype: int64
- name: character
dtype: string
- name: label
dtype: bool
- name: alice_label
dtype: bool
- name: bob_label
dtype: bool
- name: difficulty
dtype: int64
- name: difficulty_quantile
dtype: float64
splits:
- name: train
num_bytes: 26256000
num_examples: 384000
- name: validation
num_bytes: 547000
num_examples: 8000
- name: test
num_bytes: 547000
num_examples: 8000
download_size: 13465330
dataset_size: 27350000
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*