Closed mzur closed 7 years ago
Thank you so much for spending time and reading it. Basically, train_op is a crucial graph tensor which must be run by the TensorFlow session to perform a training step. As an example, performing a training step can be equivalent to updating the weights. In the code that you mentioned. train_op is the optimizer object, not just the loss.
You can refer to TensorFlow Mechanics for further details. Moreover, a detailed explanation is provided in Convolutional Neural Networks documentation.
Thanks for the explanation.
Is defining
train_op
for each data point and epoch anew really needed? I'm new to TensorFlow so I can't tell why or why not this would make sense. For me, the regression seems to work fine (and much faster) if the line is removed.