Question

我目前因我的神经网络而出现错误。数据集由2418个数组组成，每个数组包含400个数字。形状为（2418,400）。经过反复试验后，我将其重塑为（2418,400,1），因为我认为这样对模型更好。我在运行时收到的错误是：

Traceback (most recent call last):

  File "<ipython-input-15-370efbb6619c>", line 9, in <module>
    model.add(LSTM(128, input_shape=(Xreshaped), activation='relu', return_sequences=True))

  File "C:\Users\Tijev\Anaconda3\envs\tfp3.6\lib\site-packages\tensorflow\python\training\checkpointable\base.py", line 442, in _method_wrapper
    method(self, *args, **kwargs)

  File "C:\Users\Tijev\Anaconda3\envs\tfp3.6\lib\site-packages\tensorflow\python\keras\engine\sequential.py", line 160, in add
    name=layer.name + '_input')

  File "C:\Users\Tijev\Anaconda3\envs\tfp3.6\lib\site-packages\tensorflow\python\keras\engine\input_layer.py", line 231, in Input
    input_tensor=tensor)

  File "C:\Users\Tijev\Anaconda3\envs\tfp3.6\lib\site-packages\tensorflow\python\keras\engine\input_layer.py", line 107, in __init__
    name=self.name)

  File "C:\Users\Tijev\Anaconda3\envs\tfp3.6\lib\site-packages\tensorflow\python\keras\backend.py", line 876, in placeholder
    x = array_ops.placeholder(dtype, shape=shape, name=name)

  File "C:\Users\Tijev\Anaconda3\envs\tfp3.6\lib\site-packages\tensorflow\python\ops\array_ops.py", line 2077, in placeholder
    return gen_array_ops.placeholder(dtype=dtype, shape=shape, name=name)

  File "C:\Users\Tijev\Anaconda3\envs\tfp3.6\lib\site-packages\tensorflow\python\ops\gen_array_ops.py", line 5788, in placeholder
    shape = _execute.make_shape(shape, "shape")

  File "C:\Users\Tijev\Anaconda3\envs\tfp3.6\lib\site-packages\tensorflow\python\eager\execute.py", line 147, in make_shape
    raise TypeError("Error converting %s to a TensorShape: %s." % (arg_name, e))

TypeError: Error converting shape to a TensorShape: only size-1 arrays can be converted to Python scalars.

我在Stackoverflow上也看到过类似的帖子，但是在这种情况下答案没有解决。我的代码如下：

model = Sequential()
model.add(LSTM(128, input_shape=(Xreshaped), activation='relu', return_sequences=True))
model.add(Dropout(0,2))
model.add(LSTM(128, activation='relu'))
model.add(Dropout(0,2))
model.add(Dense(32, activation='relu'))
model.add(Dropout(0,2))
model.add(Dense(2, activation='sigmoid'))

opt = tf.keras.optimizers.Adam(lr=1e-3, decay=1e-5)

# mean_squared_error = mse
model.compile(loss='sparse_categorical_crossentropy',
              optimizer=opt,
              metrics=['accuracy'])

model.fit(X, Y, epochs=3, validation_data=(X, Y))

Answer 1

似乎问题已通过使用解决：

Xreshaped = X.reshape((2418,400,1))

input_shape = (400,1)

TypeError：将形状转换为TensorShape时出错：仅接受大小为1的数组

1 个答案: