Skip to content
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.

Sign up for GitHub

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

Jump to bottom

Start to use AdaDelta as optimizer #4

Closed
davidsandberg opened this issue Feb 12, 2016 · 3 comments
Closed

Start to use AdaDelta as optimizer #4

davidsandberg opened this issue Feb 12, 2016 · 3 comments

Comments

@davidsandberg
Copy link
Owner

Start to use AdaDelta as optimizer when it's available in tensorflow
tensorflow/tensorflow#644
@davidsandberg
Copy link
Owner Author

Currently AdaGrad with a learning rate lr=0.1 is used. This seems to work fairly well without putting any effort into learning rate optimization.

@davidsandberg
Copy link
Owner Author

The possibility to use Adadelta optimizer has been added. With learning_rate=0.1, rho=0.9, epsilon=1e-6 the training gets significantly slower. Need to look more into this.

@davidsandberg
Copy link
Owner Author

Did another attempt with the AdaDelta optimizer. It seems to work fairly well but no notable difference to the AdaGrad optimizer. More details can be found on the wiki: https://github.com/davidsandberg/facenet/wiki/Performance-optimization

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
1 participant
@davidsandberg