metadata
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
dataset_info:
features:
- name: task
dtype: string
- name: prompt
dtype: string
- name: correct_code
dtype: string
- name: prompt_code
dtype: string
- name: full
dtype: string
- name: index
dtype: int64
- name: noise_code
dtype: string
- name: noise_correct_code
dtype: string
splits:
- name: train
num_bytes: 3738893
num_examples: 1000
download_size: 1701993
dataset_size: 3738893