metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: embd
sequence: float64
splits:
- name: train
num_bytes: 8982174
num_examples: 541
download_size: 1739246
dataset_size: 8982174
configs:
- config_name: default
data_files:
- split: train
path: data/train-*