Search found 259 matches

by jorose
Mon Jun 10, 2019 4:32 pm
Forum: Computer Chess Club: Tournaments and Matches
Topic: Leela Chess Zero 42518 vs Stockfish 100619
Replies: 6
Views: 650

Re: Leela Chess Zero 42518 vs Stockfish 100619

I don't think you will actually be able to judge the effect it is having without a reasonably large sample size. Even if Leela has a reasonably low overhead in terms of CPU usage the effect on cache is harder to measure. It's also unclear what you would compare the performance to, perhaps one engine...
by jorose
Mon Jun 10, 2019 3:52 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: is LVA as in MVV-LVA useless ?
Replies: 9
Views: 2092

Re: is LVA as in MVV-LVA useless ?

I recently tried using my full blown move ordering (just without SEE) in QSearch, but it scored something like -8.5 +- 6 at the end of my test. I would imagine the reason was simply how expensive the calculation is. I am currently intending to train a set of weights on only root QS (as in the first ...
by jorose
Mon Jun 10, 2019 3:31 pm
Forum: Computer Chess Club: Tournaments and Matches
Topic: Leela Chess Zero 42518 vs Stockfish 100619
Replies: 6
Views: 650

Re: Leela Chess Zero 42518 vs Stockfish 100619

Considering SF play on CPU and Lz0 on GPU. It probably make sense to play with permanent brain turned on. I tried on my system couple of days ago. As long as one core is left for the OS and 1-2 core for Lz0. Have you tried this? He only has 16 physical cores so how many threads would you give Leela...
by jorose
Sat Jun 08, 2019 3:45 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Is this hard ?
Replies: 13
Views: 2404

Re: Is this hard ?

Somehow Winter has one of the two correct moves for all depths, but Im not sure at what point Winter really understood it. position fen 4B3/4kP2/8/6P1/6KP/8/8/5r2 w - - 1 67 go infinite info depth 1 seldepth 2 time 0 nodes 18 nps 18000 score cp 176 pv h4h5 info depth 2 seldepth 4 time 0 nodes 88 nps...
by jorose
Fri Jun 07, 2019 5:23 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Using evaluation hash score as current bestscore (initial guess)
Replies: 13
Views: 2478

Re: Using evaluation hash score as current bestscore (initial guess)

I think it is clear the adjusted value could technically utilize tighter bounds. How much tighter depends on the information gained from hash. On the other hand, a position found in your hash may be a more difficult one on average, so perhaps less tight bounds would be better after all. Most likely ...
by jorose
Thu Jun 06, 2019 9:25 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Using evaluation hash score as current bestscore (initial guess)
Replies: 13
Views: 2478

Re: Using evaluation hash score as current bestscore (initial guess)

If I read Xiphos code correctly then the bestscore there is just the current best score estimate for the purposes of things like futility pruning. The value gets reset before the move loop is reached. In an ideal world the static score would be equivalent to the true outcome of the game. Instead we ...
by jorose
Tue Jun 04, 2019 3:30 pm
Forum: Computer Chess Club: Tournaments and Matches
Topic: LC0 is have trouble beating Bikjump Match
Replies: 6
Views: 1072

Re: LC0 is have trouble beating Bikjump Match

I'm not sure what the point of this match was, the guy's comment was clearly regarding performance on his Android phone. BikJump is a 2100 CCRL engine so either Leela is terrible on android, which is possible given that it might be running on a phone's CPU, or it might be that he misconfigured it. L...
by jorose
Sun Jun 02, 2019 6:35 am
Forum: Computer Chess Club: General Topics
Topic: New engine releases 2019
Replies: 384
Views: 97268

Re: New engine releases 2019

I thought I would also post results of Winter 0.6 gauntlet for 4 cores. These matches have large error bars at such a small number of games, especially since it's self play. The result against the two older versions is motivatingly good though. Rank Name Elo +/- Games Score Draws 0 Winter 4c v0.6 15...
by jorose
Sat Jun 01, 2019 5:12 pm
Forum: Computer Chess Club: General Topics
Topic: New engine releases 2019
Replies: 384
Views: 97268

Re: New engine releases 2019

I ran a gauntlet against previous Winter versions at roughly CCRL 4/40 TC (though in cutechess the moves and time seem to be flipped, so the TC command was 40/240) and the results against the previous releases were roughly as expected. If there is an issue I am thinking it is either compiler or arch...
by jorose
Fri May 31, 2019 11:47 pm
Forum: Computer Chess Club: General Topics
Topic: New engine releases 2019
Replies: 384
Views: 97268

Re: New engine releases 2019

I am assuming this is on 1 core at 4/40 time control? How is the speed compared to vs 0.5? I assume there were no losses on time? What GUI are you using? Yes, it is on 1 core and the time control is 40/2 (adjusted to the reference machine). The speed was abot 836 knps for 0.5 and 788 knps for 0.6. ...