metadata
dataset_info:
features:
- name: task_type
dtype: string
- name: problem
dtype: string
- name: answer
dtype: string
- name: problem_tokens
dtype: int64
- name: answer_tokens
dtype: int64
splits:
- name: train
num_bytes: 16558492.175080558
num_examples: 11260
- name: eval
num_bytes: 446076.5
num_examples: 256
download_size: 1883637
dataset_size: 17004568.67508056
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: eval
path: data/eval-*