I think if Giraffe were converted to Cuda, we might all get our eyes opened.
Oh, wait. They already did that.
Moderators: hgm, Rebel, chrisw
I think if Giraffe were converted to Cuda, we might all get our eyes opened.
Dann Corbit wrote: ↑Fri Jun 01, 2018 7:32 amI think if Giraffe were converted to Cuda, we might all get our eyes opened.
Oh, wait. They already did that.
Ofc it would, if those 32 cores were ARM cores running at 800MHzNay Lin Tun wrote: ↑Fri Jun 01, 2018 9:11 am If Leela were running in A0 hardware, Leela would be beating stockfish 8 on 32 cores. (64 thread)
Yeah, I see the exact same thing with Leela. But I guess with more training she will overcome that as well anyway.yanquis1972 wrote: ↑Fri Jun 01, 2018 1:02 am been watching a 1'1 match between leela & SF8 on one core, & it's transparently clear that, while tactics are an obvious issue, leela's woeful misevaluation of endgames is bleeding a lot of wins (& draws). i don't really understand how, after 15 million games, it could be so bad, but she has repeatedly dissolved into drawn endgames she thinks are +2 or +3. KRB vs KR is the last example off the top of my ahead, & i think the game eval was something like +2.50.
is there any concrete evidence this is improving?
Well, for this to be a provable issue (compared to other top engines) you would need to show that other engines did better in the exact same time. Share a position where she made a mistake, entering the wrong endgame, and that SF or Komodo find the solution in that same time control.yanquis1972 wrote: ↑Fri Jun 01, 2018 1:02 am been watching a 1'1 match between leela & SF8 on one core, & it's transparently clear that, while tactics are an obvious issue, leela's woeful misevaluation of endgames is bleeding a lot of wins (& draws). i don't really understand how, after 15 million games, it could be so bad, but she has repeatedly dissolved into drawn endgames she thinks are +2 or +3. KRB vs KR is the last example off the top of my ahead, & i think the game eval was something like +2.50.
is there any concrete evidence this is improving?
Lol, here is a test for you. Play against your strongest LC0 10s per move or even more if you wish. You with naked king and give LC0 bishop and knight. You can try couple of times. And please tell us the result.Albert Silver wrote: ↑Fri Jun 01, 2018 5:01 pmWell, for this to be a provable issue (compared to other top engines) you would need to show that other engines did better in the exact same time. Share a position where she made a mistake, entering the wrong endgame, and that SF or Komodo find the solution in that same time control.yanquis1972 wrote: ↑Fri Jun 01, 2018 1:02 am been watching a 1'1 match between leela & SF8 on one core, & it's transparently clear that, while tactics are an obvious issue, leela's woeful misevaluation of endgames is bleeding a lot of wins (& draws). i don't really understand how, after 15 million games, it could be so bad, but she has repeatedly dissolved into drawn endgames she thinks are +2 or +3. KRB vs KR is the last example off the top of my ahead, & i think the game eval was something like +2.50.
is there any concrete evidence this is improving?
You are aware that KRB vs KR is quite winnable, and depends very much on the defender having the right position and not making any mistakes even then? There are excellent practical chances to win it.
No need. It used to have trouble until it learned how to win them. All this has been tracked in detail:Milos wrote: ↑Fri Jun 01, 2018 5:15 pmLol, here is a test for you. Play against your strongest LC0 10s per move or even more if you wish. You with naked king and give LC0 bishop and knight. You can try couple of times. And please tell us the result.Albert Silver wrote: ↑Fri Jun 01, 2018 5:01 pmWell, for this to be a provable issue (compared to other top engines) you would need to show that other engines did better in the exact same time. Share a position where she made a mistake, entering the wrong endgame, and that SF or Komodo find the solution in that same time control.yanquis1972 wrote: ↑Fri Jun 01, 2018 1:02 am been watching a 1'1 match between leela & SF8 on one core, & it's transparently clear that, while tactics are an obvious issue, leela's woeful misevaluation of endgames is bleeding a lot of wins (& draws). i don't really understand how, after 15 million games, it could be so bad, but she has repeatedly dissolved into drawn endgames she thinks are +2 or +3. KRB vs KR is the last example off the top of my ahead, & i think the game eval was something like +2.50.
is there any concrete evidence this is improving?
You are aware that KRB vs KR is quite winnable, and depends very much on the defender having the right position and not making any mistakes even then? There are excellent practical chances to win it.
Than take any below 2500 Elo engine without TBs and repeat. And tell us the result.
do you know who can do it in least amount of moves stockfish or lc0?Albert Silver wrote: ↑Fri Jun 01, 2018 5:41 pmNo need. It used to have trouble until it learned how to win them. All this has been tracked in detail:Milos wrote: ↑Fri Jun 01, 2018 5:15 pmLol, here is a test for you. Play against your strongest LC0 10s per move or even more if you wish. You with naked king and give LC0 bishop and knight. You can try couple of times. And please tell us the result.Albert Silver wrote: ↑Fri Jun 01, 2018 5:01 pm
Well, for this to be a provable issue (compared to other top engines) you would need to show that other engines did better in the exact same time. Share a position where she made a mistake, entering the wrong endgame, and that SF or Komodo find the solution in that same time control.
You are aware that KRB vs KR is quite winnable, and depends very much on the defender having the right position and not making any mistakes even then? There are excellent practical chances to win it.
Than take any below 2500 Elo engine without TBs and repeat. And tell us the result.
https://docs.google.com/spreadsheets/d/ ... edit#gid=0