metadata
dataset_info:
features:
- name: act
dtype: string
- name: prompt
dtype: string
- name: embedding
sequence: float32
splits:
- name: train
num_bytes: 310201
num_examples: 153
download_size: 398381
dataset_size: 310201
configs:
- config_name: default
data_files:
- split: train
path: data/train-*