chore: add model names

src/model/alexnet.py

@@ -41,7 +41,7 @@ class AlexNetModel(Sequential):
             layers.Dropout(rate=0.5),
 
             layers.Dense(name='unfreeze', units=classes, activation='softmax')
-        ])
+        ], name='alexnet')
 
     def compile(self,
                 optimizer=tf.optimizers.SGD(learning_rate=0.001),
@@ -54,7 +54,7 @@ class AlexNetModel(Sequential):
         return super().fit(x=x, y=y, batch_size=batch_size, epochs=epochs, callbacks=callbacks, **kwargs)
 
     def get_embedding_model(self):
-        core = Model(inputs=self.input, outputs=self.layers[-2].output)
+        core = Model(inputs=self.input, outputs=self.layers[-2].output, name='emb_' + self.name)
         for layer in core.layers: layer.trainable = False
         return core
 

src/model/efficientnet.py

@@ -13,7 +13,7 @@ class EfficientNetModel(Sequential):
     def __init__(self):
         super(EfficientNetModel, self).__init__([
             hub.KerasLayer(MODEL_URL, trainable=False)  # EfficientNet V2 S backbone, frozen weights
-        ])
+        ], name='efficientnet')
         self.build((None,) + TARGET_SHAPE + (3,))
 
     def compile(self, metrics=['accuracy'], **kwargs):

src/model/mobilenet.py

@@ -24,7 +24,7 @@ class MobileNetModel(Model):
         x = layers.Activation(activation='softmax', name='predictions')(x)
 
         x = tf.keras.applications.mobilenet.preprocess_input(x)
-        super(MobileNetModel, self).__init__(inputs=core.input, outputs=x)
+        super(MobileNetModel, self).__init__(inputs=core.input, outputs=x, name='mobilenet')
 
     def compile(self,
                 optimizer=tf.keras.optimizers.RMSprop(),
@@ -38,7 +38,7 @@ class MobileNetModel(Model):
 
     def get_embedding_model(self):
         core = Model(inputs=self.input, outputs=self.layers[-7].output)
-        core = Sequential([core, layers.Flatten()])
+        core = Sequential([core, layers.Flatten()], name='emb_' + self.name)
         for layer in core.layers: layer.trainable = False
         return core
 

src/model/siamese.py

@@ -55,7 +55,7 @@ class SiameseModel(Model):
         computed_distance = layers.Lambda(cosine_distance)([v1, v2])
         # computed_distance = layers.Lambda(euclidean_distance)([v1, v2])
 
-        super(SiameseModel, self).__init__(inputs=[emb_input_1, emb_input_2], outputs=computed_distance)
+        super(SiameseModel, self).__init__(inputs=[emb_input_1, emb_input_2], outputs=computed_distance, name='siamese')
 
     def get_projection_model(self):
         """ Projection model is a model from embeddings to image vector """

src/model/vgg16.py

@@ -38,7 +38,7 @@ class VGG16Model(Model):
                 **kwargs
             )
 
-        super(VGG16Model, self).__init__(inputs=model.input, outputs=model.output)
+        super(VGG16Model, self).__init__(inputs=model.input, outputs=model.output, name='vgg16')
 
     def compile(self,
                 optimizer=tf.keras.optimizers.RMSprop(),
@@ -51,7 +51,7 @@ class VGG16Model(Model):
         return super().fit(x=x, y=y, batch_size=batch_size, epochs=epochs, callbacks=callbacks, **kwargs)
 
     def get_embedding_model(self):
-        core = Model(inputs=self.input, outputs=self.layers[-2].output)
+        core = Model(inputs=self.input, outputs=self.layers[-2].output, name='emb_' + self.name)
         for layer in core.layers: layer.trainable = False
         return core
 

src/model/vit.py

@@ -12,8 +12,8 @@ MODEL_URL = "https://tfhub.dev/sayakpaul/vit_s16_fe/1"
 class VitModel(Sequential):
     def __init__(self):
         super(VitModel, self).__init__([
-            hub.KerasLayer(MODEL_URL, trainable=False)  # EfficientNet V2 S backbone, frozen weights
-        ])
+            hub.KerasLayer(MODEL_URL, trainable=False)
+        ], name='vit')
         self.build((None,) + TARGET_SHAPE + (3,))
 
     def compile(self, metrics=['accuracy'], **kwargs):