metadata
dataset_info:
features:
- name: id
dtype: string
- name: setting_key
dtype: string
- name: difficulty
dtype: string
- name: level
dtype: int32
- name: messages
sequence: string
- name: ground_truth
dtype: string
- name: dataset
dtype: string
- name: omega_split
dtype: string
splits:
- name: train
num_bytes: 222497465
num_examples: 425232
download_size: 34167171
dataset_size: 222497465
configs:
- config_name: default
data_files:
- split: train
path: data/train-*