import joblib
import json
   
with open('data/processed/go_terms.json') as f:
    go_data = json.load(f)
   
from sklearn.preprocessing import MultiLabelBinarizer
mlb = MultiLabelBinarizer()
mlb.fit([list(go_data['go_to_idx'].keys())])
joblib.dump(mlb, 'mlb_public_v1.pkl')