metadata
dataset_info:
features:
- name: stmt
dtype: string
- name: sentiment
dtype: string
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
- name: labels
sequence: int64
splits:
- name: train
num_bytes: 2834650
num_examples: 4361
- name: test
num_bytes: 315250
num_examples: 485
download_size: 35919
dataset_size: 3149900
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*