我正在运行以下keras模型
input_profile = Input(shape=(23, 23, 1))
x = Conv2D(16, (3, 3), activation='relu', padding='same')(input_profile)
x = MaxPooling2D((2, 2), padding='same')(x)
x = Conv2D(8, (3, 3), activation='relu', padding='same')(x)
x = MaxPooling2D((2, 2), padding='same')(x)
x = Conv2D(8, (3, 3), activation='relu', padding='same')(x)
encoded = MaxPooling2D((2, 2), padding='same')(x)
# at this point the representation is (4, 4, 8) i.e. 128-dimensional
x = Conv2D(8, (3, 3), activation='relu', padding='same')(encoded)
x = UpSampling2D((2, 2))(x)
x = Conv2D(8, (3, 3), activation='relu', padding='same')(x)
x = UpSampling2D((2, 2))(x)
x = Conv2D(16, (3, 3), activation='relu')(x)
x = UpSampling2D((2, 2))(x)
decoded = Conv2D(1, (3, 3), activation='sigmoid', padding='same')(x)
autoencoder = Model(input_profile, decoded)
autoencoder.compile(optimizer='adadelta', loss='binary_crossentropy')
X_GTEx = np.load('GTEx_X_float64.npy')
x_train = X_GTEx
x_train = np.reshape(x_train, (5207, 23, 23, 1))
from keras.callbacks import TensorBoard
autoencoder.fit(x_train, x_train,\
epochs=50, batch_size=127,\
shuffle=True, validation_data=(x_train, x_train),\
callbacks=[TensorBoard(log_dir='/tmp/autoencoder')])
运行它会产生以下错误:
ValueError: Error when checking target: expected conv2d_7 to have shape (20, 20, 1) but got array with shape (23, 23, 1)
很明显,我没有对shape(20,20,1)
设置任何内容。我的程序怎么了
输入的每个维度的长度都是奇数。这使得
MaxPooling2D
层使用floor
操作符对张量进行向下采样在使用}
Input(shape=(23, 23, 1))
的模型中,每个MaxPooling2D
层之后张量的维数变化将为23 to 11
{然后两个
UpSampling
层将(5, 5, 1)
张量向上采样为(20, 20, 1)
,但模型需要一个与输入形状相同的张量相关问题 更多 >
编程相关推荐