metadata
dataset_info:
features:
- name: conversations
list:
- name: from
dtype: string
- name: loss
dtype: bool
- name: value
dtype: string
- name: item_id
dtype: string
splits:
- name: train
num_bytes: 22408485
num_examples: 3313
download_size: 4514406
dataset_size: 22408485
configs:
- config_name: default
data_files:
- split: train
path: data/train-*