metadata
dataset_info:
features:
- name: token_length
dtype: int64
- name: completion
dtype: string
- name: prompt_forums_story
dtype: string
splits:
- name: train
num_bytes: 4037885.817184966
num_examples: 1000
download_size: 743525
dataset_size: 4037885.817184966
configs:
- config_name: default
data_files:
- split: train
path: data/train-*