Search found 278 matches

by brianr
Sun Feb 03, 2019 11:11 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: So, how many of you are working on neural networks for chess?
Replies: 22
Views: 1245

Re: So, how many of you are working on neural networks for chess?

Given a game state S and the result of the game R in { -1.0 (loss) , 0.0 (tie), 1.0 (win) }, is the value head trained to predict R? I think so, yes. And the policy head predicts the move with the best winning chances. How the value and policy heads work together is something I need to understand b...
by brianr
Sun Feb 03, 2019 12:45 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: So, how many of you are working on neural networks for chess?
Replies: 22
Views: 1245

Re: So, how many of you are working on neural networks for chess?

Could you summarize how Leela trains off pgns? In terms of the actual net training process, Leela learns pretty much the same way with PGNs. The major practical difference is that the PGN game positions must first be converted into the input bit planes which are fed to the Leela net training code (...
by brianr
Sat Feb 02, 2019 3:11 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: So, how many of you are working on neural networks for chess?
Replies: 22
Views: 1245

Re: So, how many of you are working on neural networks for chess?

One of the main reasons I am not working on an NN based engine is that I feel I don't have the resources. I think very few people have the resources to seriously train and test a NN based engine. Reinforcement learning based NN engines are even worse in that regard. I don't understand how it matter...
by brianr
Sat Jan 26, 2019 12:04 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Elo boost and time management
Replies: 19
Views: 1142

Re: Elo boost and time management

Time management is somewhat deceptively complex. Including sophisticated pondering, "easy move [more for the benefit of humans observing]", and considering opponent time left make it a specialized and often overlooked niche within computer chess all by itself.
by brianr
Sat Jan 26, 2019 11:52 am
Forum: Computer Chess Club: General Topics
Topic: LC0 bad search and poor tactics
Replies: 45
Views: 2860

Re: LC0 bad search and poor tactics

Yes, NN engines have a distinctly different style of play, however collectively software improvements have contributed more strength than one might think. Yes, hardware gets faster and certainly helps quite a lot, but software refinements in the aggregate are a major factor. For example: null move, ...
by brianr
Sat Jan 26, 2019 11:37 am
Forum: Computer Chess Club: General Topics
Topic: Big boost in strength for LC0 by simply changing VGA card
Replies: 19
Views: 1389

Re: Big boost in strength for LC0 by simply changing VGA card

...I am doing another match now, but this time using a Silvesuite book and things are not going as well. At the moment after 60 games Leela is down by 2 games with score at +11, =36,-13. I looked at some of the openings from book and many openings had SF up by over +1 from the opening and Leela on ...
by brianr
Wed Jan 23, 2019 1:13 pm
Forum: Computer Chess Club: General Topics
Topic: Cutechess questions
Replies: 3
Views: 621

Re: Cutechess questions

That is an older version.
Newer ones here https://github.com/cutechess/cutechess/releases

Commands https://github.com/cutechess/cutechess/ ... c/help.txt

For Q1 perhaps don't use Cutechess book, but do specify an engine book.
Q2: I'll leave to experts.
by brianr
Mon Jan 21, 2019 10:41 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: A0 policy head ambiguity
Replies: 4
Views: 695

Re: A0 policy head ambiguity

AZ and Lc0 have different "all possible moves" lists and lengths, although I have not looked at the details in a while.
by brianr
Sun Jan 20, 2019 4:25 pm
Forum: Computer Chess Club: Tournaments and Matches
Topic: LC0 11248 gauntlet vs new LC0 generation
Replies: 16
Views: 2079

Re: LC0 11248 gauntlet vs new LC0 generation

FWIW: Score of ID-32585 vs ID-11248: 61 - 63 - 176 [0.497] 300 Elo difference: -2.32 +/- 25.28 At 320 games it was exactly even. Match with v20.1 net 32585 v 11248 at 0:10+0.4 with 6 piece TBs and 2 move book on a GTX-1070. Of course, 32585 is not 32458. Update/correction. I ran a couple more match...
by brianr
Fri Jan 18, 2019 1:07 pm
Forum: Computer Chess Club: Tournaments and Matches
Topic: LC0 11248 gauntlet vs new LC0 generation
Replies: 16
Views: 2079

Re: LC0 11248 gauntlet vs new LC0 generation

FWIW:

Code: Select all

Score of ID-32585 vs ID-11248: 61 - 63 - 176  [0.497] 300
Elo difference: -2.32 +/- 25.28
At 320 games it was exactly even.

Match with v20.1 net 32585 v 11248 at 0:10+0.4 with 6 piece TBs and 2 move book on a GTX-1070.

Of course, 32585 is not 32458.