-
Notifications
You must be signed in to change notification settings - Fork 391
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Difference between _fixed_layer
and _enas_layer
in cifar10/micro_child.py
#8
Comments
Hi Ben, Thanks for the questions. I'll try.
The reason why we cannot just fix Let us know if you still have more questions 😃 |
For number 2, the point was that you're using pooling w/ stride > 1 in the fixed architecture, but a combination of Makes sense about the dynamic graphs being slow. Thanks for the quick response. (And thanks for releasing the code! I've been working on a similar project for a little while, so am very excited to compare what I've done to your code.) ~ Ben |
I think it's just because we couldn't figure out how to syntactically make |
@hyhieu I am wondering if the reduction cell in If I understand it correctly, to make the previous layers consistent, this line should be layers = [layers[0], x] |
There are a number of differences between
_fixed_layer
and_enas_layer
incifar10/micro_child.py
.Are you able to give some insight on why the code works like this? It seems that when a fixed architecture is specified, the resulting model is not necessarily exactly the same as during the RL training. It seems to me like the easiest way to fix the child architecture is to have an alternate "dummy controller", that just keeps
normal_arc
andreduce_arc
fixed at the desired architecture.Thanks
Ben
The text was updated successfully, but these errors were encountered: