我正在尝试使用TensorFlow构建一个Autoencoder模型。输入数据形状：(339, 129, 32)模型架构如下：

input_data = tf.keras.layers.Input(shape=train_sample.shape[1:])

masking = keras.layers.Masking(mask_value=-100.0)(input_data)

conv = tf.keras.Sequential([
    tf.keras.layers.Conv1D(128, 3, padding='same'),
    tf.keras.layers.AveragePooling1D(3),
    tf.keras.layers.ReLU(),
    tf.keras.layers.Conv1D(64, 3, padding='same'),
    tf.keras.layers.AveragePooling1D(2),
    tf.keras.layers.ReLU(),
    tf.keras.layers.Conv1D(32, 3, padding='same'),
    tf.keras.layers.ReLU(),
    tf.keras.layers.Conv1D(16, 3, padding='same'),
    tf.keras.layers.ReLU()
])
fc = tf.keras.Sequential([
    tf.keras.layers.Flatten(),
    tf.keras.layers.Dense(256),
    tf.keras.layers.BatchNormalization(),
    tf.keras.layers.ReLU(),
    tf.keras.layers.Dense(32),
])
conv_out = conv(masking)
fc_out = fc(conv_out)

defc = tf.keras.Sequential([
    tf.keras.layers.Dense(256),
    tf.keras.layers.BatchNormalization(),
    tf.keras.layers.ReLU(),
    tf.keras.layers.Dense(576),
    tf.keras.layers.ReLU(),
    tf.keras.layers.Dense(1024),
    tf.keras.layers.ReLU()
])

dconv = tf.keras.Sequential([
    tf.keras.layers.Reshape((1024, 1)),
    tf.keras.layers.Conv1DTranspose(576, kernel_size=3, dilation_rate=2),
    tf.keras.layers.BatchNormalization(),
    tf.keras.layers.ReLU(),
    tf.keras.layers.Conv1DTranspose(384, kernel_size=4, strides=3, padding='same'),
    tf.keras.layers.BatchNormalization(),
    tf.keras.layers.ReLU(),
    tf.keras.layers.Conv1DTranspose(32, kernel_size=3, strides=3),
    tf.keras.layers.Activation("sigmoid")
])

defc_out = defc(fc_out)
dconv_out = dconv(defc_out)

字符串
但得到以下错误：
ValueError: Dimensions must be equal, but are 9252 and 129 for '{{node mean_squared_error/SquaredDifference}} = SquaredDifference[T=DT_FLOAT](model_6/sequential_21/activation_3/Sigmoid, IteratorGetNext:1)' with input shapes: [?,9252,32], [?,129,32].个
我期待着你的帮助修复模型架构。
我试着玩模型Architure，不幸的是没有工作。