-
Notifications
You must be signed in to change notification settings - Fork 20
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
Document meaning of some of the less obvious terms of the protos. #1
Comments
I indeed intended it to be the value that resign is predicated on. I guess in terms of documentation, this should be comments in the proto files? |
Yeah I like proto file comments. |
I don't understand what value-to-us root q has, since it's up to that player at that time what move to play. We only use averaging over all moves as a probabilistic estimation of what some unknown future opponent might play, as opposed to when it the player's turn right now, there's total certainty about which move will be played (putatively speaking). It should just be the best move value, which is what we currently use for resign. (IOW, agreed with both of you.) |
meson options to disable lc0 backends
Cosmetic changes to mcts (comments, variable renames)
* alternative activation encoding * simplification Co-authored-by: borg323 <[email protected]>
Specifically at least the 'value' in the chunk data format. (If we want to use it for resign analysis it has to be the value resign is predicated on, which is the value of the best move (ignoring temperature). But root Q is something some people have expressed an interest in.)
The text was updated successfully, but these errors were encountered: