在构建和编译keras顺序模型时“列表索引超出范围”

Question

I was trying to build CNN model for classifying mnist data. Thus I tried these codes, but 'list index out of range' error has been raised

I'm working with python 3.6 and tensorflow 1.12.0, Windows10, and my IDE is PyCharm.

sess = tf.Session()

mnist = tf.keras.datasets.mnist
(x_train, y_train), (x_test, y_test) = mnist.load_data()
x_train = x_train / 255.0
x_test = x_test / 255.0
y_train = sess.run(tf.one_hot(y_train, 10))
y_test = sess.run(tf.one_hot(y_test, 10))

model = tf.keras.models.Sequential([
    tf.keras.layers.Conv2D(filters=32, kernel_size=[3, 3], strides=[1, 1],
                       padding='same', activation=tf.nn.relu),
    tf.keras.layers.MaxPooling2D(pool_size=[2, 2], strides=2, padding='same'),
    tf.keras.layers.Dropout(rate=0.3),
    tf.keras.layers.Conv2D(filters=64, kernel_size=[3, 3], strides=[1, 1],
                       padding='same', activation=tf.nn.relu),
    tf.keras.layers.MaxPooling2D(pool_size=[2, 2], strides=2, padding='same'),
    tf.keras.layers.Dropout(rate=0.3),
    tf.keras.layers.Flatten(),
    tf.keras.layers.Dense(units=512, activation=tf.nn.relu),
    tf.keras.layers.Dropout(rate=0.5),
    tf.keras.layers.Dense(units=10, activation=tf.nn.softmax)
 ])

model.compile(optimizer='adam',
          loss='categorical_crossentropy',
          metrics=['accuracy'])

model.fit(x_train, y_train, epochs=15)
print("Accuracy: {}".format(model.evaluate(x_test, y_test)))

but the result was like this:

Traceback (most recent call last):
  File 
"C:/Users/wltjd/Desktop/Computer/TensorFlow/mnist_using_CNN_advanced.py", line 38, in <module>
model.fit(x_train, y_train, epochs=15)
  File "C:\Users\wltjd\Anaconda3\envs\TensorFlow\lib\site-packages\tensorflow\python\keras\engine\training.py", line 1536, in fit
validation_split=validation_split)
  File "C:\Users\wltjd\Anaconda3\envs\TensorFlow\lib\site-packages\tensorflow\python\keras\engine\training.py", line 992, in _standardize_user_data
class_weight, batch_size)
  File "C:\Users\wltjd\Anaconda3\envs\TensorFlow\lib\site-packages\tensorflow\python\keras\engine\training.py", line 1032, in _standardize_weights
self._set_inputs(x)
  File "C:\Users\wltjd\Anaconda3\envs\TensorFlow\lib\site-packages\tensorflow\python\training\checkpointable\base.py", line 474, in _method_wrapper
method(self, *args, **kwargs)
  File "C:\Users\wltjd\Anaconda3\envs\TensorFlow\lib\site-packages\tensorflow\python\keras\engine\training.py", line 1242, in _set_inputs
self.build(input_shape=input_shape)
  File "C:\Users\wltjd\Anaconda3\envs\TensorFlow\lib\site-packages\tensorflow\python\keras\engine\sequential.py", line 222, in build
layer.build(shape)
  File "C:\Users\wltjd\Anaconda3\envs\TensorFlow\lib\site-packages\tensorflow\python\keras\layers\convolutional.py", line 190, in build
self.rank + 2))
  File "C:\Users\wltjd\Anaconda3\envs\TensorFlow\lib\site-packages\tensorflow\python\ops\nn_ops.py", line 828, in __init__
input_channels_dim = input_shape[num_spatial_dims + 1]
  File "C:\Users\wltjd\Anaconda3\envs\TensorFlow\lib\site-packages\tensorflow\python\framework\tensor_shape.py", line 616, in __getitem__
return self._dims[key]
IndexError: list index out of range

Answer 1

The error IndexError: list index out of range simply means you are trying to access a location within a list that doesn't exist.

Here, you are trying to fit x_train with y_train with one of these dataframes larger than another. I would try comparing the lengths of x_train and y_train to see which one is larger than the other and then change the sizes so they both are of equal length.