diff --git a/tests/physical_models/crop/test_leaf_dynamics.py b/tests/physical_models/crop/test_leaf_dynamics.py
index 0770237..89dbb9e 100644
--- a/tests/physical_models/crop/test_leaf_dynamics.py
+++ b/tests/physical_models/crop/test_leaf_dynamics.py
@@ -43,6 +43,15 @@ def get_test_data(file_path):
     return inputs["ModelResults"], inputs["Precision"]
 
 
+def get_test_diff_leaf_model():
+    test_data_path = phy_data_folder / "test_leafdynamics_wofost72_01.yaml"
+    params, wdp, agro, external_states = prepare_engine_input(test_data_path)
+    config_path = str(phy_data_folder / "WOFOST_Leaf_Dynamics.conf")
+    return DiffLeafDynamics(params, wdp, agro, config_path, external_states)
+
+
+
+
 class DiffLeafDynamics(torch.nn.Module):
     def __init__(self, params, wdp, agro, config_path, external_states):
         super().__init__()
@@ -132,13 +141,7 @@ def test_wofost_pp_with_leaf_dynamics(self):
 
 class TestDiffLeafDynamicsTDWI:
     def test_gradients_tdwi_lai_leaf_dynamics(self):
-        # prepare model input
-        test_data_path = phy_data_folder / "test_leafdynamics_wofost72_01.yaml"
-        params, wdp, agro, external_states = prepare_engine_input(test_data_path)
-        config_path = str(phy_data_folder / "WOFOST_Leaf_Dynamics.conf")
-
-        # create a model and optimizer
-        model = DiffLeafDynamics(params, wdp, agro, config_path, external_states)
+        model = get_test_diff_leaf_model()
         tdwi = torch.nn.Parameter(torch.tensor(0.2, dtype=torch.float32))
         output = model({"TDWI": tdwi})
         lai = output[0, :, 0]
@@ -159,12 +162,7 @@ def test_gradients_tdwi_lai_leaf_dynamics(self):
 
     def test_gradients_tdwi_twlv_leaf_dynamics(self):
         # prepare model input
-        test_data_path = phy_data_folder / "test_leafdynamics_wofost72_01.yaml"
-        params, wdp, agro, external_states = prepare_engine_input(test_data_path)
-        config_path = str(phy_data_folder / "WOFOST_Leaf_Dynamics.conf")
-
-        # create a model and optimizer
-        model = DiffLeafDynamics(params, wdp, agro, config_path, external_states)
+        model = get_test_diff_leaf_model()
         tdwi = torch.nn.Parameter(torch.tensor(0.2, dtype=torch.float32))
         output = model({"TDWI": tdwi})
         twlv = output[0, :, 1]
@@ -187,12 +185,7 @@ def test_gradients_tdwi_twlv_leaf_dynamics(self):
 class TestDiffLeafDynamicsSPAN:
     def test_gradients_span_lai_leaf_dynamics(self):
         # prepare model input
-        test_data_path = phy_data_folder / "test_leafdynamics_wofost72_01.yaml"
-        params, wdp, agro, external_states = prepare_engine_input(test_data_path)
-        config_path = str(phy_data_folder / "WOFOST_Leaf_Dynamics.conf")
-
-        # create a model and optimizer
-        model = DiffLeafDynamics(params, wdp, agro, config_path, external_states)
+        model = get_test_diff_leaf_model()
         span = torch.nn.Parameter(torch.tensor(30, dtype=torch.float32))
         output = model({"SPAN": span})
         lai = output[0, :, 0]
@@ -213,12 +206,7 @@ def test_gradients_span_lai_leaf_dynamics(self):
 
     def test_gradients_span_twlv_leaf_dynamics(self):
         # prepare model input
-        test_data_path = phy_data_folder / "test_leafdynamics_wofost72_01.yaml"
-        params, wdp, agro, external_states = prepare_engine_input(test_data_path)
-        config_path = str(phy_data_folder / "WOFOST_Leaf_Dynamics.conf")
-
-        # create a model and optimizer
-        model = DiffLeafDynamics(params, wdp, agro, config_path, external_states)
+        model = get_test_diff_leaf_model()
         span = torch.nn.Parameter(torch.tensor(30, dtype=torch.float32))
         output = model({"SPAN": span})
         twlv = output[0, :, 1]