David J Wu
David J Wu
Thanks for the feedback. What is the optimal number of threads that the benchmark command concludes that you should use? And is this a setup with multiple GPUs, or only...
Thanks for posting this. I will add more variations of this kind of pincer to the positions for self-play next time I update the positions, likely within the next month...
@Arce0416 yes, this method does work so long as the shape that contains the mistake is one that is likely to reoccur. In general, if you include data about all...
This was how KataGo learned to evaluate Mi Yuting's "flying dagger" joseki more accurately - this joseki is a widespread weakness of pure "zero" AI bots because it's sharp enough...
It takes some manual work to update all the hint positions on the server, and there's a large overhead to reuploading all the positions with new weights and then recomputing...
So it would be much too soon to see anything, maybe check back in a couple of months, now that the hint position is there?
@kaorahi - Very cool. Is this phenomenon consistent across multiple different games? It would be interesting to know if one could also find other ways to try to objectively measure...
Thanks for reporting this way back, I missed it earlier but yes it is correct and I have gotten around to making the fix now as of https://github.com/lightvector/KataGo/releases/tag/v1.15.2 and master...
Thanks! As with the previous cases, can you attach an SGF?
Ah, thanks! I didn't realize this was the same example as another from before.