metadata
dataset_info:
features:
- name: question
dtype: string
- name: answer
dtype: string
- name: source
dtype: string
- name: domain
sequence: string
- name: llama8b_solve_rate
dtype: float64
splits:
- name: train
num_bytes: 9685421.014486982
num_examples: 31600
download_size: 3613125
dataset_size: 9685421.014486982
configs:
- config_name: default
data_files:
- split: train
path: data/train-*