metadata
dataset_info:
features:
- name: model_name
dtype: string
- name: metric_name
dtype: string
- name: value
dtype: float64
splits:
- name: train
num_bytes: 1084
num_examples: 28
download_size: 2151
dataset_size: 1084
configs:
- config_name: default
data_files:
- split: train
path: data/train-*