metadata
dataset_info:
features:
- name: text
dtype: string
- name: level
dtype: string
- name: topic
dtype: string
- name: embeddings
sequence: float16
splits:
- name: train
num_bytes: 1321487569
num_examples: 275725
download_size: 847331872
dataset_size: 1321487569
configs:
- config_name: default
data_files:
- split: train
path: data/train-*