metadata
dataset_info:
features:
- name: sample_idx
dtype: int64
- name: metric
dtype: string
- name: score
dtype: float64
- name: Model Nickname
dtype: string
- name: benchmark_and_optional_task
dtype: string
splits:
- name: train
num_bytes: 1044088
num_examples: 17632
download_size: 7478
dataset_size: 1044088
configs:
- config_name: default
data_files:
- split: train
path: data/train-*