metadata
dataset_info:
features:
- name: category
dtype: string
- name: source
dtype: string
- name: text
dtype: string
- name: corrupted_text
dtype: string
- name: score
dtype: float64
- name: tags
sequence: int64
splits:
- name: train
num_bytes: 65201521
num_examples: 79832
- name: test
num_bytes: 8149957
num_examples: 10000
download_size: 92193012
dataset_size: 73351478
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*