metadata
dataset_info:
features:
- name: id
dtype: string
- name: difficulty
dtype: string
- name: messages
dtype: string
- name: expected_tool_call
dtype: string
- name: rationale
dtype: string
splits:
- name: train
num_bytes: 38565
num_examples: 160
download_size: 15147
dataset_size: 38565
configs:
- config_name: default
data_files:
- split: train
path: data/train-*