Peter Szoldan questions

Peter Szoldan answers

• Is this a correct way to reuse a tf.slim graph for validation during training?
• Clip to maximum / minimum finite int/float/double?
• L2 regularization keep increasing during training
• Set "training=False" of "tf.layers.batch_normalization" when training will get a better validation result
• How to provide custom gradient in TensorFlow
• Calculate face_descriptor faster
• What is colocate_with used for in tensorflow?
• How ensure that Keras is using GPU with tensorflow backend?
• Second derivative in Keras
• Is tf.contrib.layers.fully_connected() behavior change between tensorflow 1.3 and 1.4 an issue?