metadata
dataset_info:
features:
- name: text
dtype: string
- name: conversation_id
dtype: int64
- name: embedding
sequence: float32
- name: compressibility
dtype: float64
splits:
- name: train
num_bytes: 387305305
num_examples: 19999
download_size: 368401496
dataset_size: 387305305
configs:
- config_name: default
data_files:
- split: train
path: data/train-*