metadata
dataset_info:
features:
- name: Arabic_text
dtype: string
- name: sentiment
dtype:
class_label:
names:
'0': Positive
'1': Both
'2': Neutral
'3': Negative
- name: confidence
dtype: string
splits:
- name: train
num_bytes: 140868
num_examples: 1200
download_size: 79288
dataset_size: 140868
configs:
- config_name: default
data_files:
- split: train
path: data/train-*