OmarMedhat22 / Iris-Recognition-CASIA-Iris-Thousand

33 stars 16 forks source link

DenseNet201 initialization - list indices must be integers or slices, not str #12

Open Dkwcs opened 11 months ago

Dkwcs commented 11 months ago

During the running of iris_classification_2.ipynb got an error:

TypeError Traceback (most recent call last) f:\repo\Iris-Recognition-CASIA-Iris-Thousand\iris_classification_2.ipynb Cell 7 line 8 5 from keras.applications.xception import Xception 7 #original_model = VGG16(include_top=False) ----> 8 original_model = DenseNet201(include_top="False") 9 #original_model = InceptionV3(include_top=False) 10 #original_model = InceptionResNetV2(include_top=False) 11 #original_model = Xception(include_top=False) 13 bottleneck_input = original_model.get_layer(index=0).input

File c:\Users\user\AppData\Local\Programs\Python\Python39\lib\site-packages\keras\applications\densenet.py:408, in DenseNet201(include_top, weights, input_tensor, input_shape, pooling, classes, classifier_activation) 395 @keras_export( 396 "keras.applications.densenet.DenseNet201", "keras.applications.DenseNet201" 397 ) (...) 405 classifier_activation="softmax", 406 ): 407 """Instantiates the Densenet201 architecture.""" --> 408 return DenseNet( 409 [6, 12, 48, 32], 410 include_top, 411 weights, 412 input_tensor, 413 input_shape, ... 68 def getitem(self, key): 69 """Retrieves item by the specified key.""" ---> 70 return self._values[key]

TypeError: list indices must be integers or slices, not str