metadata
dataset_info:
features:
- name: prefix
dtype: string
- name: suffix
dtype: string
- name: prefix_tokens
dtype: int64
- name: suffix_tokens
dtype: int64
- name: sample_id
dtype: string
- name: category
dtype: string
- name: is_canary
dtype: bool
- name: canary_pii_type
dtype: string
- name: canary_value
dtype: string
- name: token_offset
dtype: int64
splits:
- name: train
num_bytes: 319278
num_examples: 600
download_size: 158747
dataset_size: 319278
configs:
- config_name: default
data_files:
- split: train
path: data/train-*