fix bug with batchnorm initialization

aidge_backend_cpu/unit_tests/test_recipies.py

@@ -31,7 +31,7 @@ class test_recipies(unittest.TestCase):
 
         input_node = aidge_core.Producer(input_tensor, "X")
         conv = aidge_core.Conv2D(1, 1, [3, 3], name="Conv0")
-        bn = aidge_core.BatchNorm2D(name="Add0")
+        bn = aidge_core.BatchNorm2D(1, name="Add0")
 
         graph_view = aidge_core.sequential([conv, bn])