metadata
dataset_info:
features:
- name: env
dtype: string
- name: prompt
dtype: string
- name: extra
dtype: string
- name: timestamp
dtype: string
- name: task_type
dtype: string
splits:
- name: train
num_bytes: 360633650
num_examples: 5999
download_size: 202470209
dataset_size: 360633650
configs:
- config_name: default
data_files:
- split: train
path: data/train-*