metadata
dataset_info:
features:
- name: task_name
dtype: string
- name: score
dtype: float64
- name: success
dtype: bool
- name: time_taken
dtype: float64
- name: extra
struct:
- name: conversation
list:
- name: content
dtype: string
- name: role
dtype: string
- name: dataset_index
dtype: int64
- name: seed
dtype: int64
- name: test_result
dtype: string
- name: usage
struct:
- name: completion_tokens
dtype: int64
- name: completion_tokens_details
dtype: 'null'
- name: prompt_tokens
dtype: int64
- name: prompt_tokens_details
dtype: 'null'
- name: total_tokens
dtype: int64
- name: challenge
dtype: string
splits:
- name: train
num_bytes: 550226111
num_examples: 1923
download_size: 87733956
dataset_size: 550226111
configs:
- config_name: default
data_files:
- split: train
path: data/train-*