metadata
dataset_info:
features:
- name: source
dtype: string
- name: book_id
dtype: string
- name: question_id
dtype: string
- name: conv_id
dtype: int64
- name: round
dtype: int64
- name: messages
list:
- name: role
dtype: string
- name: content
dtype: string
- name: tool_calls
list:
- name: id
dtype: string
- name: function
struct:
- name: name
dtype: string
- name: arguments
dtype: string
- name: type
dtype: string
- name: tool_call_id
dtype: string
splits:
- name: train
num_bytes: 313730027
num_examples: 21505
download_size: 68625927
dataset_size: 313730027
configs:
- config_name: default
data_files:
- split: train
path: data/train-*