Update expected failures.

aidge_interop_torch/unit_tests/test_models.py

 import torch
 import aidge_interop_torch
 import aidge_core
+import aidge_onnx
 import aidge_learning
 import pytest
 
@@ -67,6 +68,7 @@ test_cases = [
         "input_size": (BATCH_SIZE, 1, 32, 32),
         "output_size": (BATCH_SIZE, 10),
         "test_backward": False,
+        "expected_failure": "No handling of MaxPool optional output"
     },
     {
         "name": "LeNet_train",
@@ -82,7 +84,7 @@ test_cases = [
         "input_size": (BATCH_SIZE, 3, 96, 96),
         "output_size": (BATCH_SIZE, 10),
         "test_backward": False,
-        "expected_failure": "forward() not implemented yet for operator of type AveragePool"
+        "expected_failure": "Fail to load AveragePool from onnx (no native coverage)."
     },
     {
         "name": "MobileNet_train",
@@ -90,7 +92,7 @@ test_cases = [
         "input_size": (BATCH_SIZE, 1, 96, 96),
         "output_size": (BATCH_SIZE, 10),
         "test_backward": True,
-        "expected_failure": "forward() not implemented yet for operator of type AveragePool"
+        "expected_failure": "Fail to load AveragePool from onnx (no native coverage)."
     },
     {
         "name": "Perceptron_eval",
@@ -112,7 +114,7 @@ test_cases = [
         "input_size": (BATCH_SIZE, 3, 32, 32),
         "output_size": (BATCH_SIZE, 10),
         "test_backward": False,
-        "expected_failure": "forward() not implemented yet for operator of type AveragePool"
+        "expected_failure": "Fail to load AveragePool from onnx (no native coverage)."
     },
     {
         "name": "ResNet_train",
@@ -120,7 +122,7 @@ test_cases = [
         "input_size": (BATCH_SIZE, 3, 32, 32),
         "output_size": (BATCH_SIZE, 10),
         "test_backward": True,
-        "expected_failure": "forward() not implemented yet for operator of type AveragePool"
+        "expected_failure": "Fail to load AveragePool from onnx (no native coverage)."
     },
 
 ]
@@ -141,7 +143,7 @@ def model_test_case(request):
 
     # Wrap with Aidge
     aidge_model = aidge_interop_torch.wrap(torch_model, input_size)
-
+    assert aidge_onnx.has_native_coverage(aidge_model._graph_view), f"Model has not a native coverage.\n{aidge_onnx.native_coverage_report(aidge_model._graph_view)}"
     # Optional: Set optimizer for Aidge model
     if cfg["test_backward"]:
         opt = aidge_learning.SGD()