Search found 337 matches

by brianr
Wed Jul 24, 2019 4:28 pm
Forum: Computer Chess Club: General Topics
Topic: Is the 320x24b larger net the strongest around for RTX GPU?
Replies: 63
Views: 6823

Re: Is the 320x24b larger net the strongest around for RTX GPU?

There is a clear difference in match results with and without the restart=on option in cutechess-cli. Below are two identical Leelas. Score of A vs B: 31 - 18 - 51 [0.565] 100 Elo difference: 45.42 +/- 47.89 Dead even within one game with restart=on Interesting, but not entirely clear to me. In the...
by brianr
Wed Jul 24, 2019 2:26 pm
Forum: Computer Chess Club: General Topics
Topic: Is the 320x24b larger net the strongest around for RTX GPU?
Replies: 63
Views: 6823

Re: Is the 320x24b larger net the strongest around for RTX GPU?

There is a clear difference in match results with and without the restart=on option in cutechess-cli.
Below are two identical Leelas.

Score of A vs B: 31 - 18 - 51 [0.565] 100
Elo difference: 45.42 +/- 47.89

Dead even within one game with restart=on
by brianr
Sun Jul 21, 2019 5:31 pm
Forum: Computer Chess Club: General Topics
Topic: buying a new computer
Replies: 280
Views: 40343

Re: buying a new computer

The huge memory usage of Lc0 is a returning theme on the sites of LC0. But this is the first established answer for it. Why, Dear "crem"? Is it so great secret? I can not believe the developers of Leela can not make modifiable the usage of RAM with an UCI parameter. Maybe it would decrease the Leel...
by brianr
Thu Jul 18, 2019 11:17 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: A question to MCTS + NN experts
Replies: 8
Views: 2191

Re: A question to MCTS + NN experts

The net architectures used with chess are very similar to those used in general image detection and classification, which work off of 2D grids of pixels (with additional color channels in more dimensions), so you can roughly think of the piece type 2D bitmaps like a piece list and each type of piece...
by brianr
Wed Jul 17, 2019 12:12 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: A question to MCTS + NN experts
Replies: 8
Views: 2191

Re: A question to MCTS + NN experts

I am far from an expert, but here are some suggestions. You mentioned SL NN experience, but perhaps start with something simpler than chess. This repo I found to be very helpful (there is another one for chess after understanding Connect-4; the overall process is the same) https://github.com/Zeta36/...
by brianr
Sun Jul 14, 2019 1:22 pm
Forum: Computer Chess Club: General Topics
Topic: Building a 500$ PC only for Chess
Replies: 3
Views: 525

Re: Building a 500$ PC only for Chess

A starting point:
https://www.tomshardware.com/reviews/be ... ,4390.html

Newer AMD 3000 cpus might be a better bet.
by brianr
Wed Jul 10, 2019 11:21 pm
Forum: Computer Chess Club: General Topics
Topic: For data lovers only
Replies: 27
Views: 2691

Re: For data lovers only

Are the links not the ones in the first post?
by brianr
Thu Jun 27, 2019 11:10 pm
Forum: Computer Chess Club: General Topics
Topic: Congratulations to Jon Dart
Replies: 11
Views: 2195

Re: Congratulations to Jon Dart

I would like to echo the above.
Moreover, Jon has shared source code for many years along with being an active and most helpful member of the computer chess community.
by brianr
Mon Jun 17, 2019 11:16 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Neural Net Endgame technique
Replies: 7
Views: 2648

Re: Neural Net Endgame technique

I did some lone king training with small nets in September last year. It was with a version of Leela that might still have had the 50 move draw issue. In any case, an 8x1 net was fine for KQ or KR v lone K. However, for KBNvK an 8x1 net could not mate. Needed to be at least 16x2 for about 2/3rds win...
by brianr
Sat Jun 08, 2019 9:42 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Hash collision?
Replies: 42
Views: 6485

Re: Hash collision?

Turns out that it may not matter very much:
http://www.craftychess.com/hyatt/collisions.html