metadata
dataset_info:
features:
- name: source
dtype: string
- name: emotion
dtype: string
- name: text
dtype: string
- name: input_ids
sequence: int32
- name: labels
sequence: int64
- name: attention_mask
sequence: int8
splits:
- name: train
num_bytes: 4345936486
num_examples: 397559
download_size: 1353265874
dataset_size: 4345936486
configs:
- config_name: default
data_files:
- split: train
path: data/train-*