metadata
dataset_info:
features:
- name: _id
dtype: string
- name: prompt
dtype: string
- name: answer
dtype: string
- name: base
dtype: int64
- name: a
dtype: int64
- name: b
dtype: int64
splits:
- name: train
num_bytes: 2458473
num_examples: 20000
- name: validation
num_bytes: 2418048
num_examples: 20000
download_size: 2310376
dataset_size: 4876521
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*