metadata
dataset_info:
features:
- name: hidden_states
sequence: float32
- name: labels
dtype: int64
splits:
- name: train
num_bytes: 6689568
num_examples: 408
- name: test
num_bytes: 6705964
num_examples: 409
download_size: 14637648
dataset_size: 13395532
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*