metadata
dataset_info:
features:
- name: task_id
dtype: string
- name: embeddings
sequence: float32
splits:
- name: train
num_bytes: 3529470
num_examples: 1140
download_size: 4359708
dataset_size: 3529470
configs:
- config_name: default
data_files:
- split: train
path: data/train-*