metadata
dataset_info:
features:
- name: input
dtype: string
- name: part
dtype: string
- name: diag_confidence
dtype: float64
splits:
- name: train
num_bytes: 103954
num_examples: 1200
- name: test
num_bytes: 12831
num_examples: 150
- name: eval
num_bytes: 12780
num_examples: 150
download_size: 38743
dataset_size: 129565
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: eval
path: data/eval-*