metadata
dataset_info:
features:
- name: input
sequence:
sequence:
sequence: float32
- name: output
sequence:
sequence:
sequence: float32
- name: text
sequence: int64
splits:
- name: train
num_bytes: 1579660000
num_examples: 1000
download_size: 358014726
dataset_size: 1579660000