data_feature/processed_data/init_2013/2019/split_meta.json

{
  "year": 2019,
  "mode": "adaptation",
  "split_source": "../splits/2019_split.json",
  "train_size": 63304,
  "test_size": 15827
}