Search found 296 matches

by Raphexon
Tue Sep 29, 2020 9:38 am
Forum: Computer Chess Club: General Topics
Topic: NNUE variation
Replies: 8
Views: 386

Re: NNUE variation

If anything that shows that you can just take a network, add some very minor modifications and pass the similarity test with ease.

Could be interesting to test Vondele's net since that's a Sergio net with the last layer SPSA tuned.

https://tests.stockfishchess.org/nns
by Raphexon
Tue Sep 29, 2020 9:10 am
Forum: Computer Chess Club: General Topics
Topic: A Crossroad in Computer Chess; Or Desperate Flailing for Relevance
Replies: 53
Views: 1269

Re: A Crossroad in Computer Chess; Or Desperate Flailing for Relevance

Of course, time passed and the Leela team managed to reach similar strength, but without the use of GPUs -- a total repudiation of my original stance. Can you clarify this? What is the strength of Leela-CPU? On CCRL, I see what I guess is Lc0 without GPU: Lc0 0.26.2 w703810 64-bit 3151 . Do you mea...
by Raphexon
Tue Sep 29, 2020 7:56 am
Forum: Computer Chess Club: General Topics
Topic: A Crossroad in Computer Chess; Or Desperate Flailing for Relevance
Replies: 53
Views: 1269

Re: A Crossroad in Computer Chess; Or Desperate Flailing for Relevance

Now, in recent months, a similar series of developments are happening. NNUEs, or rather, I want to be specific, the King-Piece structures used in Stockfish, are flooding the scene. With a few hours work, an author can copy paste some code from CFish or Stockfish, add the incremental update code, an...
by Raphexon
Fri Sep 25, 2020 9:58 pm
Forum: Computer Chess Club: General Topics
Topic: sf with black pieces
Replies: 7
Views: 622

Re: sf with black pieces

You shouldn't conclude too much from 55 games.
by Raphexon
Thu Sep 24, 2020 11:29 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: First success with neural nets
Replies: 17
Views: 1112

Re: First success with neural nets

Right now I. 'm using floating points with some SIMD. I estimate it could be about 3 times faster if I change to 16-bit fixed point with SIMD and add incremental updates for changing inputs on the first layer. After optimization I might play with net size and inputs to see if what works best. If yo...
by Raphexon
Fri Sep 18, 2020 9:23 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Off-the-shelf PC spec, engine development
Replies: 34
Views: 1348

Re: Off-the-shelf PC spec, engine development

Money now vs time later.
The 3990x will let you test twice as fast.
by Raphexon
Tue Sep 15, 2020 7:34 am
Forum: Computer Chess Club: General Topics
Topic: Chess variants made with help from alpha zero article
Replies: 37
Views: 2026

Re: Chess variants made with help from alpha zero article

The start position of Capablanca is relatively garbage. Testing it with Fairy-Stockfish gave me a fairly hefty opening advantage at 60+0.6. White vs Black: +952 -438 =610 [0.628] 2000 There are also unprotected pawns in the beginning. I've also tested Gothic Chess. (Capablanca but different start p...
by Raphexon
Tue Sep 15, 2020 7:28 am
Forum: Computer Chess Club: Tournaments and Matches
Topic: SPCC: Testrun of Ethereal 12.50 finished
Replies: 1
Views: 290

Re: SPCC: Testrun of Ethereal 12.50 finished

Congratz to Andrew, that's a hefty upgrade over 12.25.
by Raphexon
Mon Sep 14, 2020 4:49 pm
Forum: Computer Chess Club: General Topics
Topic: Chess variants made with help from alpha zero article
Replies: 37
Views: 2026

Re: Chess variants made with help from alpha zero article

The start position of Capablanca is relatively garbage. Testing it with Fairy-Stockfish gave me a fairly hefty opening advantage at 60+0.6. White vs Black: +952 -438 =610 [0.628] 2000 There are also unprotected pawns in the beginning. I've also tested Gothic Chess. (Capablanca but different start p...
by Raphexon
Mon Sep 14, 2020 2:21 pm
Forum: Computer Chess Club: General Topics
Topic: Chess variants made with help from alpha zero article
Replies: 37
Views: 2026

Re: Chess variants made with help from alpha zero article

The start position of Capablanca is relatively garbage. Testing it with Fairy-Stockfish gave me a fairly hefty opening advantage at 60+0.6. White vs Black: +952 -438 =610 [0.628] 2000 There are also unprotected pawns in the beginning. I've also tested Gothic Chess. (Capablanca but different start p...