metadata
dataset_info:
features:
- name: conversations
list:
- name: from
dtype: string
- name: value
dtype: string
- name: value_ru
dtype: string
- name: weight
dtype: float64
- name: key
dtype: int64
splits:
- name: train
num_bytes: 183635644
num_examples: 47536
download_size: 83293621
dataset_size: 183635644
configs:
- config_name: default
data_files:
- split: train
path: data/train-*