如何在使用Tensorflow数据集API时为标量提供设置

Question

我使用TF Dataset API和一个占位符作为文件名，在初始化迭代器时输入这些文件名（不同的文件取决于它是训练集还是验证集）。我还想使用额外的占位符来指示我们是在培训还是在验证（包括在退出层中）。但是，我无法使用数据集初始值设定项将值馈送到此占位符（这是有意义的，因为它不是数据集的一部分）。如何在使用数据集API时提供额外的变量呢？你知道吗

关键代码：

filenames_placeholder = tf.placeholder(tf.string, shape = (None))
is_training = tf.placeholder(tf.bool, shape = ()) # Error: You must feed a value for placeholder tensor 'Placeholder_1' with dtype bool
dataset = tf.data.TFRecordDataset(filenames_placeholder)
# (...) Many other dataset operations
iterator = dataset.make_initializable_iterator()
next_element = iterator.get_next()

# Model code using "next_element"  as inputs including the dropout layer at some point 
# where I would like to let the model know if we're training or validating

tf.layers.dropout(x, training = is_training)

# Model execution
with tf.Session() as sess:
sess.run(tf.global_variables_initializer())
sess.run(iterator.initializer, feed_dict = {filenames_placeholder: training_files, is_training: True})
# (...) Performing training
sess.run(iterator.initializer, feed_dict = {filenames_placeholder: training_files, is_training: False})
# (...) Performing validadtion

Answer 1

在这种情况下，我要做的是使用一个默认值创建一个额外的占位符：

keep_prob = tf.placeholder_with_default(1.0, shape=())

在图表中：

tf.layers.dropout(inputs, rate=1-keep_prob)

培训时：

sess.run(...,feed_dict={keep_prob:0.5})

评估时：

sess.run(...) # No feed_dict here since the keep_prob placeholder has a default value of 1

注意，在训练时输入一个占位符，它提供了一个额外的float值，并不会降低训练的速度。你知道吗