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.

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

Question regarding architecture #147

Open
GRIGORR opened this issue Feb 10, 2020 · 3 comments
Open

Question regarding architecture #147

GRIGORR opened this issue Feb 10, 2020 · 3 comments
Labels
question Further information is requested

Comments

@GRIGORR
Copy link

GRIGORR commented Feb 10, 2020

Hi. Small question regarding the architecture. Do I understand correctly that at each iteration the candidate networks are trained separately, then each of them is added to Adanet, weighting the logits is learned and then the best one is chosen? Thanks in advance.

@le-dawg
Copy link

le-dawg commented Feb 26, 2020

I understan it like that too.

@cweill
Copy link
Contributor

cweill commented Apr 7, 2020

@GRIGORR That is correct: all the candidate subnetworks (and their associated ensemble) are trained in parallel in the same TensorFlow graph. At the end of each iteration, the best subnetwork is chosen based on its performance within the ensemble.

@cweill cweill added the question Further information is requested label Apr 7, 2020
@le-dawg
Copy link

le-dawg commented Apr 7, 2020

@GRIGORR That is correct: all the candidate subnetworks (and their associated ensemble) are trained in parallel in the same TensorFlow graph. At the end of each iteration, the best subnetwork is chosen based on its performance within the ensemble.

From what I gather from the 0.8.0 docs it sounds to me like one AdaNet iteration actually selects a complete Ensemble each iteration and discards the others. Could it be said that each Ensemble from the candidate ensemble set differs from all the other candidate Ensembles in the subnetwork that has been added to it in the current iteration?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
question Further information is requested
Projects
None yet
Development

No branches or pull requests

3 participants