metadata
dataset_info:
features:
- name: text
dtype: string
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
- name: logits
sequence:
sequence: float64
splits:
- name: train
num_bytes: 121553662
num_examples: 1
download_size: 92949005
dataset_size: 121553662
configs:
- config_name: default
data_files:
- split: train
path: data/train-*