metadata
dataset_info:
features:
- name: subset
dtype: string
- name: model
dtype: string
- name: conversation
list:
- name: content
dtype: string
- name: role
dtype: string
- name: annotations
list:
- name: index
dtype: int64
- name: label
dtype: string
- name: span
dtype: string
- name: verification_note
dtype: string
- name: canary
dtype: string
splits:
- name: train
num_bytes: 4696337
num_examples: 492
- name: validation
num_bytes: 589779
num_examples: 59
- name: test
num_bytes: 9159152
num_examples: 996
download_size: 6825086
dataset_size: 14445268
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*