The early stopping Plugin parameterization #1382
Unanswered
AymenTlili131
asked this question in
Q&A
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
Hi , I'm trying to train multiple models each taking a lengthy amount of time if they were to go through all the epoch . I'd like them to stop when convergence is achieved (I'll describe this as not improving by more than 0.05% in terms of accuracy)
so to me accuracy at epoch e: 95.00% and epoch e+1: 95.04% are the same and I'd like to exit at epoch e+1 while also checkpointing .
wanted to ask if the following parametrization is correct when working on 2 experiments of the splitMnist Benchmark .
for a more wholistic view and code to test with (note the if statement limiting this to one init/activation combination
Beta Was this translation helpful? Give feedback.
All reactions