metadata
dataset_info:
features:
- name: tweet
dtype: string
- name: offensive
dtype:
class_label:
names:
'0': NOT_OFF
'1': 'OFF'
- name: hate
dtype:
class_label:
names:
'0': NOT_HS
'1': HS
splits:
- name: train
num_bytes: 1390414
num_examples: 6838
- name: validation
num_bytes: 200610
num_examples: 999
download_size: 822093
dataset_size: 1591024
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*