TalkChess.com

Thanks. Fab got TRASHED I have done a little bit of work on FabChess in the last month. If I get in some more work in the next month, there might be enough changes for a new release before the next season of TCEC.

But how do I do this at depth 2? - First generate moves at depth 1 and store them in a queue with d=1, giving 20 moves... - Then generate all moves at depth 2 and store them in the same queue with d=2... giving 400 moves and the original 20. You don't store the moves in a queue but the resulting po...

@mar: Thanks for reminding me of the parallel for. Rust has the "rayon" crate which can do that; but AFAIK, it tries to use all physical cores. I'll see what rayon can do. It has been some time since Ive looked into it. Executing the threading from ply two is a good idea, because it gives you a lot...

You can get away with 16 bits, for storing the entire move information? Interesting. (Or did you mean 16 bytes? But that would be a lot.) I already have these: PIECE : 3 0-7 (use only 0-6) FROM : 6 0-63 TO : 6 0-63 CAPTURE : 3 0-7 (captured piece) PROMOTION : 3 0-7 (piece promoted to) ENPASSANT : 1...

You can get away with 16 bits, for storing the entire move information? Interesting. (Or did you mean 16 bytes? But that would be a lot.) I already have these: PIECE : 3 0-7 (use only 0-6) FROM : 6 0-63 TO : 6 0-63 CAPTURE : 3 0-7 (captured piece) PROMOTION : 3 0-7 (piece promoted to) ENPASSANT : 1...

All this fancy stuff will put you in a straitjacket, I don't know about Rust ('Rust roest' we say here in Holland), but I still think C/C++ is the only serious language to write a chess engine in. Most other languages output slower code (or the compilers are worse), for instance I keep a shadow cop...

According to FabianVDW (of FabChess, which is also an engine in Rust), mutex locking works, but is too slow if an engine uses more than 8 threads. Then the overhead of the locks/unlocks becomes so big that more threads don't add any more speed. He says that "racey access to the hash table is necess...

Ethereal, Weiss, and now Fabchess have found large elo gains tuning using these datasets. Weiss was able to implement the tuner a la the paper and found +50 elo in 60s+.6s with the initial tune. I believe Weiss was already tuning using an older dataset of mine, and with his own python framework. An...

I can chime in on two of the points and maybe provide a bit of a mathematical background, although I am unsure how much of this you know already. 1) If I want the NN to return an evaluation, which could be used in a traditional alphabeta search engine, how is this achieved most commonly? Do I have a...

Thanks for this tourney Graham! Fab overperformed pretty hard, but still almost managed to throw the lead away in the end