I'm planing to use ResNet to do a classification task, and find tensorflow implementation here, howerev, there are something I don't understand as a newbie to tensorflow.
while not mon_sess.should_stop(): mon_sess.run(model.train_op)
how it kowns when to stop training?
I will very appreciate it if there are some explaination.
The batching is decided in the input reading code. The training loop is the code you pasted (while not should_stop, run train op). It will stop training when the input has been exhausted, according to the same input processing code.