metadata
dataset_info:
features:
- name: input_ids
sequence: int32
- name: prompt_text
dtype: string
splits:
- name: train
num_bytes: 5044435
num_examples: 4096
download_size: 2659014
dataset_size: 5044435
configs:
- config_name: default
data_files:
- split: train
path: data/train-*