metadata
dataset_info:
features:
- name: conversations
list:
- name: from
dtype: string
- name: value
dtype: string
splits:
- name: train
num_bytes: 2204869925
num_examples: 3530293
- name: validation
num_bytes: 182520693
num_examples: 268953
- name: test
num_bytes: 204631665
num_examples: 270998
download_size: 1193685160
dataset_size: 2592022283
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*