4
$\begingroup$

If I'm doing a hyperparameter search and comparing two different hyperparameters (but not number of epochs), is there some established rule of thumb for how many epochs to run? If I just compare after a few epochs, will that give me a good idea about how it will perform fully converged (say for example after 1000 epochs). Is there any research papers where people have studied this question?

$\endgroup$
1
$\begingroup$

The general rule of thumb is to run the number of epochs until validation error starts to increase.

Sometimes fast initial learning will not lead to the best performance later.

| improve this answer | |
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.