Python CNN LSTM（值误差步长应为 1、1 或 3 但实际为 2）

Question

我一直在尝试在 mnist 数据集上训练一个 convlstm 模型，以拓宽我在模型开发方面的知识。我无法逃避我在标题中包含的错误。任何帮助或提示表示赞赏！

我知道步幅的默认值是 (1,1) 但不确定 2 是如何设置的。

import tensorflow as tf

from keras.models import Sequential

from keras.layers import Dense, Dropout, LSTM, CuDNNLSTM, TimeDistributed, Reshape

from keras.utils import to_categorical

from keras.layers.convolutional import Conv2D, Conv3D

from keras.layers.pooling import MaxPooling2D, MaxPool3D

from keras.layers.core import Flatten

def prep_pixels(train, test):

    # convert from integers to floats

    train_norm = train.astype('float32')

    test_norm = test.astype('float32')

    # normalize to range 0-1

    train_norm = train_norm / 255.0

    test_norm = test_norm / 255.0

    # return normalized images

    return train_norm, test_norm

mnist = tf.keras.datasets.mnist

(x_train, y_train), (x_test, y_test) = mnist.load_data()

x_train = x_train.reshape((x_train.shape[0], 28, 28, 1))

x_test = x_test.reshape((x_test.shape[0], 28, 28, 1))

y_train = to_categorical(y_train)

y_test = to_categorical(y_test)

x_train, x_test = prep_pixels(x_train, x_test)

model = Sequential()

model.add(TimeDistributed(Conv2D(32, (3, 3), activation='relu', input_shape=(28, 28, 1))))

model.add(TimeDistributed((MaxPooling2D((2, 2)))))

model.add(TimeDistributed(Flatten()))

model.add(LSTM(32, activation='relu', return_sequences=True))

model.add(Dropout(0.2))

model.add(Dense(10, activation='softmax'))

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

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

model.fit(x_train, y_train, epochs=1, validation_data=(x_test, y_test))

错误

model.fit(x_train, y_train, epochs=1, validation_data=(x_test, y_test))

strides = _get_sequence(strides, n, channel_index, "strides")

ValueError：步幅应该是长度 1、1 或 3 但实际是 2

Answer 1

您似乎还没有为 ConvLSTM 创建窗口数据集。所以你可能想在打电话之前这样做model.fitd_train = tf.keras.preprocessing.sequence.TimeseriesGenerator(x_train, y_train, length=5, batch_size=64) # window size = 5d_test = tf.keras.preprocessing.sequence.TimeseriesGenerator(x_test, y_test, length=5)model.fit(d_train, epochs=1, validation_data=d_test)为了与您的损失函数保持一致，您需要禁用返回序列（或添加另一个不返回序列的层）。model.add(tf.keras.layers.LSTM(32, activation='relu', return_sequences=False))