metadata
license: apache-2.0
dataset_info:
features:
- name: prompt
dtype: string
- name: text
dtype: string
- name: source
dtype: string
- name: label
dtype: int64
- name: perplexity
dtype: float64
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
splits:
- name: train
num_bytes: 1244922869
num_examples: 330345
- name: test
num_bytes: 65575515
num_examples: 17387
download_size: 435690981
dataset_size: 1310498384
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*