metadata
dataset_info:
features:
- name: original_text
dtype: string
- name: neighbors
list:
- name: log_prob
dtype: float64
- name: text
dtype: string
splits:
- name: train
num_bytes: 143489307
num_examples: 5000
download_size: 14292841
dataset_size: 143489307
configs:
- config_name: default
data_files:
- split: train
path: data/train-*