metadata
dataset_info:
features:
- name: id
dtype: string
- name: sample_idx
dtype: int32
- name: turn_idx
dtype: int32
- name: cp_type
dtype: string
- name: prefix
dtype: json
- name: tools
dtype: json
- name: expected_tool
dtype: string
- name: expected_args
dtype: json
- name: expected_content
dtype: string
- name: expected_response
dtype: string
splits:
- name: train
num_bytes: 1082581
num_examples: 85
download_size: 127108
dataset_size: 1082581
configs:
- config_name: default
data_files:
- split: train
path: data/train-*