metadata
dataset_info:
features:
- name: answer
dtype: string
- name: length
dtype: int64
- name: prompt
dtype: string
- name: conversations
list:
- name: content
dtype: string
- name: role
dtype: string
splits:
- name: train
num_bytes: 2033641715.4176462
num_examples: 494678
download_size: 1199928106
dataset_size: 2033641715.4176462
configs:
- config_name: default
data_files:
- split: train
path: data/train-*