metadata
dataset_info:
features:
- name: sample_type
dtype: string
- name: model_input
dtype: string
- name: response
dtype: string
- name: text
dtype: string
splits:
- name: train
num_bytes: 2498946549.1212187
num_examples: 669081
- name: test
num_bytes: 131524288.87878108
num_examples: 35215
download_size: 1834868352
dataset_size: 2630470838
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*