metadata
dataset_info:
features:
- name: task_id
dtype: large_string
- name: source
dtype: large_string
- name: train_pairs
dtype: large_string
- name: test_pairs
dtype: large_string
splits:
- name: train
num_bytes: 4686553
num_examples: 1000
download_size: 816568
dataset_size: 4686553
configs:
- config_name: default
data_files:
- split: train
path: data/train-*