metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: completion
dtype: string
- name: numbers
list: int64
- name: answer
dtype: int64
splits:
- name: train
num_bytes: 2995142
num_examples: 9000
- name: validation
num_bytes: 332851
num_examples: 1000
download_size: 727927
dataset_size: 3327993
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*