TalkChess.com

A few questions: 1) the nnue should be ~80% of the speed of hce of something like Scorpio (so 20% slower). After the compiler optimizations Daniel suggested, are you seeing something similar? 2) nnue is updated a little bit with each move/null move (and undo of a move). That’s what makes it efficien...

Now this looks interesting! Let lc0 switch to another NN during the game (PR 573 rebased to current master) #1452 https://github.com/LeelaChessZero/lc0/pull/1452 Wonder what NN is best at endgames? DKappe are you listening? :D Medium Ender 192x16, which Scorpio uses but is already over a year old. ...

Btw, if I'm now using SF NNUE does it mean that I need to change the licence to GNU GPL 3 like in SF? If all of the code has licenses that you can change, that would be easiest, but if you distribute MIT and GPL code together, it’s really more of an issue for someone wanting to do something that th...

Suppose that LC0 were trained not from self-play games but from data extracted from its games on TCEC, CCRL, and CEGT? Since those contests include lots of arbitrary openings, many of which are imbalanced, I wonder if it would help the problem. 1. Too little data. 2. There have been a number of exp...

[1] Do people really want to see a bunch of NNUE engines, all running the same evaluation function, occupying the top slots of computer chess? Prove it. The Toga III nets, trained off of Toga II 4.0.1, play nothing like SF NNUE. It’s surprising to see so many smart technical people think of neural ...

Could you please explain it to me then? If I'm missing crucial information, I'm perfectly happy to change my view when presented with additional information which brings my current view into question. I’m happy to explain things to someone who listens and discusses in good faith. Your posts here an...

Well, the DeusX thing is one of the weirder things in TCEC history. Its a major reason for the concern over people lying about origins of their NNUEs. Since we've already shown that one can attempt to lie, get called out, but still compete, only to be removed after the fact. No one was impressed by...

When I played in Leiden in December 2018 (and won) I was very forthcoming about using the GPLed Lc0 binary to run my homebrewed NN "DeusX". I explained what had gone into it: human games, engine games, selfplay games, tablebases, etc. to set it apart from Leela's net That is a nice change of pace f...

Yes! Your nets are great. Which would you say is best for a 2200 player like me to train against? Which is influenced more by the human liChess games? BTW...was very close to supporting you on Patreaon the other day, but got distracted by my kids. Love your work...still doing it? No need to support...

HOWEVER ... the 3rd best move, 12. Qe2 lets you keep a small edge with best play from your opponent while giving your opponent far more chances to go wrong . This idea of 'wrong' would need to be within a reasonable 'human evaluation' of those replies. In other words it would not count a queen drop...