metadata
dataset_info:
features:
- name: data_name
dtype: string
- name: task_name
dtype: string
- name: exp_id
dtype: string
- name: llm_log
dtype: string
- name: results
dtype: string
- name: scores
list: float64
splits:
- name: train
num_bytes: 21603089471
num_examples: 23200
download_size: 8518050367
dataset_size: 21603089471
configs:
- config_name: default
data_files:
- split: train
path: data/train-*