Move ResNet ImageNet downloading in separate function.

keras_retinanet/models/resnet.py

@@ -26,6 +26,28 @@
 custom_objects.update(keras_resnet.custom_objects)
 
 
+def download_imagenet(backbone):
+    allowed_backbones = [50, 101, 152]
+    if backbone not in allowed_backbones:
+        raise ValueError('Backbone (\'{}\') not in allowed backbones ({}).'.format(backbone, allowed_backbones))
+
+    filename = resnet_filename.format(backbone)
+    resource = resnet_resource.format(backbone)
+    if backbone == 50:
+        checksum = '3e9f4e4f77bbe2c9bec13b53ee1c2319'
+    elif backbone == 101:
+        checksum = '05dc86924389e5b401a9ea0348a3213c'
+    elif backbone == 152:
+        checksum = '6ee11ef2b135592f8031058820bb9e71'
+
+    return keras.applications.imagenet_utils.get_file(
+        filename,
+        resource,
+        cache_subdir='models',
+        md5_hash=checksum
+    )
+
+
 def resnet_retinanet(num_classes, backbone=50, inputs=None, weights='imagenet', **kwargs):
     allowed_backbones = [50, 101, 152]
     if backbone not in allowed_backbones:
@@ -37,21 +59,7 @@ def resnet_retinanet(num_classes, backbone=50, inputs=None, weights='imagenet',
 
     # determine which weights to load
     if weights == 'imagenet':
-        filename = resnet_filename.format(backbone)
-        resource = resnet_resource.format(backbone)
-        if backbone == 50:
-            checksum = '3e9f4e4f77bbe2c9bec13b53ee1c2319'
-        elif backbone == 101:
-            checksum = '05dc86924389e5b401a9ea0348a3213c'
-        elif backbone == 152:
-            checksum = '6ee11ef2b135592f8031058820bb9e71'
-
-        weights_path = keras.applications.imagenet_utils.get_file(
-            filename,
-            resource,
-            cache_subdir='models',
-            md5_hash=checksum
-        )
+        weights_path = download_imagenet(backbone)
     elif weights is None:
         weights_path = None
     else: