metadata
dataset_info:
features:
- name: org_dataset
dtype: string
- name: uniq_id
dtype: string
- name: en_messages
list:
- name: content
dtype: string
- name: role
dtype: string
- name: hi_messages
list:
- name: content
dtype: string
- name: role
dtype: string
splits:
- name: train
num_bytes: 24849043
num_examples: 9498
download_size: 12293374
dataset_size: 24849043
configs:
- config_name: default
data_files:
- split: train
path: data/train-*