Question

我在使用tensorflow恢复模型时使用批处理规范有一点问题。

以下是我的批量规范，来自here：

def _batch_normalization(self, input_tensor, is_training, batch_norm_epsilon, decay=0.999):
    """batch normalization for dense nets.

    Args:
        input_tensor: `tensor`, the input tensor which needed normalized.
        is_training: `bool`, if true than update the mean/variance using moving average,
                             else using the store mean/variance.
        batch_norm_epsilon: `float`, param for batch normalization.
        decay: `float`, param for update move average, default is 0.999.

    Returns:
        normalized params.
    """
    # actually batch normalization is according to the channels dimension.
    input_shape_channels = int(input_tensor.get_shape()[-1])

    # scala and beta using in the the formula like that: scala * (x - E(x))/sqrt(var(x)) + beta
    scale = tf.Variable(tf.ones([input_shape_channels]))
    beta = tf.Variable(tf.zeros([input_shape_channels]))

    # global mean and var are the mean and var that after moving averaged.
    global_mean = tf.Variable(tf.zeros([input_shape_channels]), trainable=False)
    global_var = tf.Variable(tf.ones([input_shape_channels]), trainable=False)

    # if training, then update the mean and var, else using the trained mean/var directly.
    if is_training:
        # batch norm in the channel axis.
        axis = list(range(len(input_tensor.get_shape()) - 1))
        batch_mean, batch_var = tf.nn.moments(input_tensor, axes=axis)

        # update the mean and var.
        train_mean = tf.assign(global_mean, global_mean * decay + batch_mean * (1 - decay))
        train_var = tf.assign(global_var, global_var * decay + batch_var * (1 - decay))
        with tf.control_dependencies([train_mean, train_var]):
            return tf.nn.batch_normalization(input_tensor,
                                             batch_mean, batch_var, beta, scale, batch_norm_epsilon)
    else:
        return tf.nn.batch_normalization(input_tensor,
                                         global_mean, global_var, beta, scale, batch_norm_epsilon)

我训练模型并使用tf.train.Saver()保存它。以下是测试代码：

def inference(self, images_for_predict):
    """load the pre-trained model and do the inference.

    Args:
        images_for_predict: `tensor`, images for predict using the pre-trained model.

    Returns:
        the predict labels.
    """

    tf.reset_default_graph()
    images, labels, _, _, prediction, accuracy, saver = self._build_graph(1, False)

    predictions = []
    correct = 0
    with tf.Session() as sess:
        sess.run(tf.global_variables_initializer())
        # saver = tf.train.import_meta_graph('./models/dense_nets_model/dense_nets.ckpt.meta')
        # saver.restore(sess, tf.train.latest_checkpoint('./models/dense_nets_model/'))
        saver.restore(sess, './models/dense_nets_model/dense_nets.ckpt')
        for i in range(100):
            pred, corr = sess.run([tf.argmax(prediction, 1), accuracy],
                                  feed_dict={
                                      images: [images_for_predict.images[i]],
                                      labels: [images_for_predict.labels[i]]})
            correct += corr
            predictions.append(pred[0])
    print("PREDICTIONS:", predictions)
    print("ACCURACY:", correct / 100)

但预测结果总是非常糟糕，如：

('PREDICTIONS:', [2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2])

('ACCURACY:', 0.080000000000000002)

一些提示：images_for_predict = mnist.test和self._build_graph方法有两个参数：batch_size和is_training。

任何人都可以帮助我？

Answer 1

在尝试了很多方法之后，我解决了这个问题，下面就是我所做的。

首先感谢@gdelab，我改为使用Doh! There was a problem，所以我的批处理规范函数就是这样：

tf.layers.batch_normalization

param def _batch_normalization(self, input_tensor, is_training): return tf.layers.batch_normalization(input_tensor, training=is_training)是一个占位符：is_training

在构建图表时，请记住在优化中添加此代码：

is_training = tf.placeholder(tf.bool)

因为extra_update_ops = tf.get_collection(tf.GraphKeys.UPDATE_OPS) with tf.control_dependencies(extra_update_ops): train_step = tf.train.AdamOptimizer(self.learning_rate).minimize(cross_entropy)添加更新均值和方差不会自动添加为列车操作的依赖关系 - 所以如果你不做任何额外的事情，他们永远不会被运行。

因此，在完成培训后，开始训练网络，使用类似的代码保存模型：

tf.layers.batch_normalization

请注意saver = tf.train.Saver(var_list=tf.global_variables()) savepath = saver.save(sess, 'here_is_your_personal_model_path')参数确保张量流保存所有参数包括设置为无法训练的全局均值/ var。

恢复并测试模型时，请执行以下操作：

var_list=tf.global_variables()

现在可以测试他/她的模型，希望它可以帮助你，谢谢！

Answer 2

看到批处理规范的实现，当您加载模型时，需要使用images, labels, _, _, prediction, accuracy, saver = self._build_graph(1, False)构建图表并为chekpoint加载权重值，但 NOT 元图。我认为saver.restore(sess, './models/dense_nets_model/dense_nets.ckpt')现在也恢复元图（抱歉，如果我错了），所以你只需要恢复它的“数据”部分。

否则，您只是使用图表进行培训，其中批次规范中使用的均值和方差是从批次中获得的。但是，当您测试批次的大小为1时，通过批次的均值和方差进行归一化总是会使您的数据为0，因此输出不变。

在任何情况下，我建议您使用tf.layers.batch_normalization代替is_training占位符，以便将其提供给您的网络...

在恢复模型时使用批量规范？

2 个答案: