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

Question about C_SCALE value. #67

Open
CGLemon opened this issue Jun 13, 2023 · 2 comments
Open

Question about C_SCALE value. #67

CGLemon opened this issue Jun 13, 2023 · 2 comments
Labels
question Further information is requested

Comments

@CGLemon
Copy link
Contributor

CGLemon commented Jun 13, 2023

According to paper, the best C_SCALE value is 0.1 on 9x9 board. Why did you use 1. Did you tune it by your personal experiment?
@kobanium
Copy link
Owner

The paper says "Experimentally, cvisit = 50, cscale = 1.0 allowed us to use the same hyperparameters
even if changing the number of simulations.". I think you mentioned about Figure 5(b), it is the result of the beam_rider.
Needless to say, C_SCALE = 1.0 might be not optimal value, it brings reasonably good results on Go.

@kobanium kobanium added the question Further information is requested label Jun 14, 2023
@CGLemon
Copy link
Contributor Author

CGLemon commented Jun 14, 2023

Thanks!

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

No branches or pull requests
2 participants
@kobanium @CGLemon