Update the DALI EfficientNet example to be compatible with the latest NumPy (#5561)

- removes deprecated usage of some NumPy types in favor of the build-ones
  (np.int->int)

Signed-off-by: Janusz Lisiecki <[email protected]>

Author: JanuszL

docs/examples/use_cases/pytorch/efficientnet/image_classification/autoaugment.py

@@ -75,7 +75,7 @@ def __init__(self):
             "translateY": np.linspace(0, 250, 11),
             "rotate": np.linspace(0, 30, 11),
             "color": np.linspace(0.1, 1.9, 11),
-            "posterize": np.round(np.linspace(0, 4, 11), 0).astype(np.int),
+            "posterize": np.round(np.linspace(0, 4, 11), 0).astype(int),
             "solarize": np.linspace(0, 256, 11),
             "solarizeadd": np.linspace(0, 110, 11),
             "contrast": np.linspace(0.1, 1.9, 11),