metadata
dataset_info:
features:
- name: session_id
dtype: string
- name: role
dtype: string
- name: content
dtype: string
- name: ts
dtype: string
- name: sources_json
dtype: string
- name: web_search
dtype: bool
- name: model
dtype: string
splits:
- name: train
num_bytes: 157767
num_examples: 51
download_size: 71597
dataset_size: 157767
configs:
- config_name: default
data_files:
- split: train
path: data/train-*