metadata
dataset_info:
features:
- name: speak
dtype: string
- name: write
dtype: string
splits:
- name: train
num_bytes: 2887027702
num_examples: 848161
download_size: 1453821474
dataset_size: 2887027702
configs:
- config_name: default
data_files:
- split: train
path: data/train-*