metadata
dataset_info:
features:
- name: transformed_prompt
dtype: string
- name: transformed_completion
dtype: string
splits:
- name: train
num_bytes: 92506680
num_examples: 550635
download_size: 853095
dataset_size: 92506680
configs:
- config_name: default
data_files:
- split: train
path: data/train-*