Merge pull request #68 from ChEB-AI/fix/retrieve_splits_from_csv

load processed data correctly when retrieving splits from csv

Author: sfluegel05

chebai/preprocessing/datasets/base.py

@@ -1045,9 +1045,7 @@ def _retrieve_splits_from_csv(self) -> None:
         splits_df = pd.read_csv(self.splits_file_path)
 
         filename = self.processed_file_names_dict["data"]
-        data = torch.load(
-            os.path.join(self.processed_dir, filename), weights_only=False
-        )
+        data = self.load_processed_data(filename=filename)
         df_data = pd.DataFrame(data)
 
         train_ids = splits_df[splits_df["split"] == "train"]["id"]