metadata
dataset_info:
features:
- name: task
dtype: string
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
- name: labels
sequence: int64
- name: raw_inputs
dtype: string
- name: raw_labels
dtype: string
splits:
- name: train
num_bytes: 453956162
num_examples: 130319
- name: validation
num_bytes: 21741236
num_examples: 5937
- name: test
num_bytes: 21676416
num_examples: 5936
download_size: 68057559
dataset_size: 497373814
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*