metadata
dataset_info:
features:
- name: id
dtype: string
- name: contribution
dtype: string
- name: recipe
dtype: string
- name: contributions_embedding
sequence: float64
splits:
- name: train
num_bytes: 437570983
num_examples: 16026
- name: test_standard_impact
num_bytes: 40368060
num_examples: 1472
- name: test_high_impact
num_bytes: 4633811
num_examples: 169
download_size: 461807056
dataset_size: 482572854
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test_standard_impact
path: data/test_standard_impact-*
- split: test_high_impact
path: data/test_high_impact-*