metadata
dataset_info:
features:
- name: text
dtype: string
- name: embeddings_A
sequence: float32
- name: embeddings_B
sequence: float32
splits:
- name: train
num_bytes: 1054505434
num_examples: 100000
download_size: 1013356324
dataset_size: 1054505434
configs:
- config_name: default
data_files:
- split: train
path: data/train-*