metadata
dataset_info:
features:
- name: query
dtype: string
- name: completion
dtype: string
- name: horizon
dtype: int64
- name: num_tokens
dtype: int64
splits:
- name: train
num_bytes: 2726412416.2004204
num_examples: 1917534
- name: test
num_bytes: 302935660.79957986
num_examples: 213060
download_size: 392808821
dataset_size: 3029348077
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*