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

Why use RESET_AFTER during training? #1

Open
littlelid opened this issue Mar 27, 2017 · 2 comments
Open

Why use RESET_AFTER during training? #1

littlelid opened this issue Mar 27, 2017 · 2 comments

Comments

@littlelid
Copy link

No description provided.

@littlelid littlelid changed the title Why use RESET_AFTER in training? Why use RESET_AFTER during training? Mar 27, 2017
@littlelid
Copy link
Author

Because RAM is enough for RESET_AFTER steps to build graph?

@bogatyy
Copy link
Owner

bogatyy commented Mar 28, 2017

Yes. Since every per-tree graph is created from scratch in RAM and kept there, we need to destroy all the past graphs from RAM every once in a while.

I arbitrarily set RESET_AFTER=50, that constant is not really particularly optimized for anything, you just can't have it too large (too small doesn't seem to hurt)

Repository owner deleted a comment from rnoh96 Mar 1, 2024
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
3 participants
@bogatyy @littlelid and others