metadata
dataset_info:
features:
- name: image
dtype: image
- name: text
dtype: string
- name: Noise_level
dtype: int64
splits:
- name: train
num_bytes: 962292803.875
num_examples: 91977
download_size: 963706057
dataset_size: 962292803.875
configs:
- config_name: default
data_files:
- split: train
path: data/train-*