diff --git a/src/stamp/modeling/data.py b/src/stamp/modeling/data.py
index c070c24..3604ce8 100755
--- a/src/stamp/modeling/data.py
+++ b/src/stamp/modeling/data.py
@@ -132,7 +132,6 @@ def tile_bag_dataloader(
         collate_fn=collate_fn,
         worker_init_fn=Seed.get_loader_worker_init() if Seed._is_set() else None,
         persistent_workers=(num_workers > 0),
-        pin_memory=torch.cuda.is_available(),
     )
 
     return (
@@ -416,7 +415,6 @@ def create_dataloader(
             num_workers=num_workers,
             worker_init_fn=Seed.get_loader_worker_init() if Seed._is_set() else None,
             persistent_workers=(num_workers > 0),
-            pin_memory=torch.cuda.is_available(),
         )
         return dl, categories or []
     else:
diff --git a/src/stamp/modeling/deploy.py b/src/stamp/modeling/deploy.py
index 035769f..6d81a8f 100644
--- a/src/stamp/modeling/deploy.py
+++ b/src/stamp/modeling/deploy.py
@@ -342,6 +342,8 @@ def deploy_categorical_model_(
                 patient_label=patient_label,
                 ground_truth_label=ground_truth_label,
                 cut_off=cut_off,
+                time_label=time_label,
+                status_label=status_label,
             ).to_csv(output_dir / f"patient-preds-{model_i}.csv", index=False)
         else:
             df_builder(
@@ -351,6 +353,8 @@ def deploy_categorical_model_(
                 patient_label=patient_label,
                 ground_truth_label=ground_truth_label,
                 cut_off=cut_off,
+                time_label=time_label,
+                status_label=status_label,
             ).to_csv(output_dir / "patient-preds.csv", index=False)
 
     if task == "classification":
@@ -641,6 +645,8 @@ def _to_survival_prediction_df(
     ],
     predictions: Mapping[PatientId, torch.Tensor],
     patient_label: PandasLabel,
+    time_label: PandasLabel = "time",
+    status_label: PandasLabel = "event",
     cut_off: float | None = None,
     **kwargs,
 ) -> pd.DataFrame:
@@ -671,9 +677,9 @@ def _to_survival_prediction_df(
         # call .split on ground-truth values — assume structured input. If
         # the value is not a 2-tuple/list, treat both fields as unknown.
         if isinstance(gt, (tuple, list)) and len(gt) == 2:
-            row["time"], row["event"] = gt
+            row[time_label], row[status_label] = gt
         else:
-            row["time"], row["event"] = None, None
+            row[time_label], row[status_label] = None, None
 
         rows.append(row)
 
diff --git a/src/stamp/preprocessing/__init__.py b/src/stamp/preprocessing/__init__.py
index b2daa38..95564b9 100755
--- a/src/stamp/preprocessing/__init__.py
+++ b/src/stamp/preprocessing/__init__.py
@@ -311,14 +311,12 @@ def extract_(
                 default_slide_mpp=default_slide_mpp,
             )
             # Parallelism is implemented in the dataset iterator already, so one worker is enough!
-            # pin_memory speeds up CPU→GPU DMA for tile batches.
             # num_workers=1 is intentional: WSI read parallelism is inside _supertiles.
             dl = DataLoader(
                 ds,
                 batch_size=64,
                 num_workers=1,
                 drop_last=False,
-                pin_memory=torch.cuda.is_available(),
             )
 
             feats, xs_um, ys_um = [], [], []