diff --git a/tests/__pycache__/__init__.cpython-38.pyc b/tests/__pycache__/__init__.cpython-38.pyc
new file mode 100644
index 0000000..6be8d5c
Binary files /dev/null and b/tests/__pycache__/__init__.cpython-38.pyc differ
diff --git a/tests/__pycache__/test_dice.cpython-38-pytest-6.2.4.pyc b/tests/__pycache__/test_dice.cpython-38-pytest-6.2.4.pyc
index 853fdf4..c7ad467 100644
Binary files a/tests/__pycache__/test_dice.cpython-38-pytest-6.2.4.pyc and b/tests/__pycache__/test_dice.cpython-38-pytest-6.2.4.pyc differ
diff --git a/tests/__pycache__/test_meshnet.cpython-38-pytest-6.2.4.pyc b/tests/__pycache__/test_meshnet.cpython-38-pytest-6.2.4.pyc
new file mode 100644
index 0000000..b97588e
Binary files /dev/null and b/tests/__pycache__/test_meshnet.cpython-38-pytest-6.2.4.pyc differ
diff --git a/tests/test_meshnet.py b/tests/test_meshnet.py
new file mode 100644
index 0000000..af24076
--- /dev/null
+++ b/tests/test_meshnet.py
@@ -0,0 +1,34 @@
+import os
+import torch
+import pytest
+from app.code.executor.meshnet import enMesh_checkpoint
+
+def test_modelAE_from_production_config():
+    # Emulate production: get the modelAE.json from app/code/executor/
+    # Import the package to use its __file__ attribute.
+    import app.code.executor  
+    config_file_path = os.path.join(os.path.dirname(app.code.executor.__file__), "modelAE.json")
+    
+    # Instantiate the model using production parameters.
+    model = enMesh_checkpoint(
+        in_channels=1,
+        n_classes=3,
+        channels=5,
+        config_file=config_file_path
+    )
+    
+    # Create a small random input tensor.
+    x = torch.randn(1, 1, 16, 16, 16)
+    
+    # Test in train mode (which uses checkpoint_sequential)
+    model.train()
+    output_train = model(x)
+    expected_shape = (1, 3, 16, 16, 16)  # Expecting n_classes=3 as output channels.
+    assert output_train.shape == expected_shape, f"Train mode output shape mismatch: expected {expected_shape}, got {output_train.shape}"
+    
+    # Test in eval mode (using inference mode)
+    model.eval()
+    with torch.no_grad():
+        output_eval = model(x)
+    assert output_eval.shape == expected_shape, f"Eval mode output shape mismatch: expected {expected_shape}, got {output_eval.shape}"
+