diff --git a/code/datasets/simplicial_ds.py b/code/datasets/simplicial_ds.py
index 80ce31e..d701489 100644
--- a/code/datasets/simplicial_ds.py
+++ b/code/datasets/simplicial_ds.py
@@ -49,17 +49,14 @@ def __init__(
         pre_transform=None,
         pre_filter=None,
     ):
-
+        self.manifold = manifold
         self.task_type = task_type
         self.split = mode
         self.split_config = SplitConfig(split, seed, use_stratified)
-        is_full = pre_filter is None
         self.raw_simplicial_ds = SimplicialDataset(
             os.path.join(
                 root,
-                "raw_simplicial",
-                f"manifold_{manifold}",
-                f"is_full_{is_full}",
+                "raw_simplicial"
             ),
             manifold,
             version,
@@ -108,12 +105,18 @@ def process(self):
         indices = range(self.raw_simplicial_ds.len())
 
         for task_type in TaskType:
+            
+            # no name classification on 3 manifolds
+            if self.manifold == "3" and task_type == TaskType.NAME:
+                continue
+
             # apply class transform
             class_transform = Compose(class_transforms_lookup[task_type])
             data_list_processed = [
                 class_transform(self.raw_simplicial_ds.get(idx))
                 for idx in indices
             ]
+            
             # train test split
             stratified = torch.vstack([data.y for data in data_list_processed])
             train_val_indices, test_indices = train_test_split(
diff --git a/code/datasets/transforms.py b/code/datasets/transforms.py
index 9ccfd94..3ab726e 100644
--- a/code/datasets/transforms.py
+++ b/code/datasets/transforms.py
@@ -20,6 +20,7 @@ def __call__(self, data):
 
 class OrientableToClassTransform:
     def __call__(self, data):
+        print(data)
         data.y = data.orientable.long()
         return data