metadata
dataset_info:
features:
- name: inputs
sequence: float64
- name: position_ids
sequence: int64
- name: token_type_ids
sequence: int8
- name: strands
dtype: int64
- name: labels
dtype: float64
splits:
- name: train
num_bytes: 476110293
num_examples: 28620
- name: val
num_bytes: 73602173
num_examples: 3948
download_size: 48052201
dataset_size: 549712466
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: val
path: data/val-*