Add custom callbacks to model training

Add an optional parameter for calling a list of keras.callbacks to be add to the original list.

mrcnn/model.py

@@ -2272,7 +2272,7 @@ def set_log_dir(self, model_path=None):
             "*epoch*", "{epoch:04d}")
 
     def train(self, train_dataset, val_dataset, learning_rate, epochs, layers,
-              augmentation=None):
+              augmentation=None, custom_callbacks=[]):
         """Train the model.
         train_dataset, val_dataset: Training and validation Dataset objects.
         learning_rate: The learning rate to train with
@@ -2299,6 +2299,10 @@ def train(self, train_dataset, val_dataset, learning_rate, epochs, layers,
                     imgaug.augmenters.Fliplr(0.5),
                     imgaug.augmenters.GaussianBlur(sigma=(0.0, 5.0))
                 ])
+	custom_callbacks: (list) Optional. Add custom callbacks to be called
+	    with the keras fit_generator method. Must be list of type keras.callbacks. 
+
+                
         """
         assert self.mode == "training", "Create model in training mode."
 
@@ -2330,6 +2334,9 @@ def train(self, train_dataset, val_dataset, learning_rate, epochs, layers,
             keras.callbacks.ModelCheckpoint(self.checkpoint_path,
                                             verbose=0, save_weights_only=True),
         ]
+
+        # Add custom callbacks to the list
+        callbacks+=custom_callbacks
 
         # Train
         log("\nStarting at epoch {}. LR={}\n".format(self.epoch, learning_rate))