Final Release of Ethereal, V12.75

[Hrvoje Horvatic]

https://github.com/AndyGrant/Ethereal/tree/nnue
Pushed NNUE support to a branch. You can load in a file. Stockfish weights are preferred.
If anyone tests this for a rating list, it should be called "Ethereal 12.75 SF-NNUE".
That branch actually works, unlike the Etherlito crap which broke due to memory alignments on POSIX.

YEAAAAAAHHHHHH!

[Hrvoje Horvatic]

SF doesn't have a closed team of developers. Dozens of people have contributed to it, and the NNUE authors are among them.

Perhaps we have a different understanding of what that means.

Apparently we have, even though what I wrote is not very complex.

I guess in your view 99% of Stockfish is not an achievement of Stockfish. That's fine, I just don't think it is a very useful view. Stockfish is not the effort of a closed team of contributors.

Btw, if I am not mistaken, the original Shogi NNUE engine was based on Stockfish.

+1

[cdani]

I remember when Daniel Jose Queralto (I hope I've spelled that correctly) said he was ceasing development of Andscacs. He said it was because there was too little progress for a large time investment. I didn't think for a minute he meant progress against Stockfish or Ethereal. He meant progress of Andscacs dev against the official release version 0.95. That's all. His measure seemed, at least to me, internal i.e. Andscacs dev vs Andscacs 0.95, not a case of Andscacs never winning TCEC or topping a ratings list.

Yes, I meant progress of Andscacs itself. Of course is nice to improve also against others, but the motivation was the one said.

[Rebel]

https://github.com/AndyGrant/Ethereal/tree/nnue
Pushed NNUE support to a branch. You can load in a file. Stockfish weights are preferred.
If anyone tests this for a rating list, it should be called "Ethereal 12.75 SF-NNUE".
That branch actually works, unlike the Etherlito crap which broke due to memory alignments on POSIX.

Anyone who is willing to share an x64-popcount version?

[Albert Silver]

SF doesn't have a closed team of developers. Dozens of people have contributed to it, and the NNUE authors are among them.

Perhaps we have a different understanding of what that means.

Apparently we have, even though what I wrote is not very complex.

I guess in your view 99% of Stockfish is not an achievement of Stockfish. That's fine, I just don't think it is a very useful view. Stockfish is not the effort of a closed team of contributors.

Btw, if I am not mistaken, the original Shogi NNUE engine was based on Stockfish.

Apparently what I wrote, which was also not complex, is not understood at all since literally nothing what you wrote above corresponds to what I stated in any way or form. And yes, you are mistaken about the Shogi NNUE engine. There are many engines in Shogi, almost all of which have NNUE support (all that I know of, but maybe not all), and they are not a list of 'SF clones'. They borrowed ideas, not copy-and-pasted code.

http://yaneuraou.yaneu.com/2020/06/19/s ... ete-guide/

"For the last five years or so, the developers of the shogi AI have borrowed a lot of ideas from Stockfish."

In fact, the actual NNUE codebase in SF that was added is of complex status concerning the GPL. I speak specifically of the learning/training code. Sure, it is licensed under GPL, and I am not arguing that, but who's GPL license is the real question. There is a reason the code is full of internal comments that bear no resemblance to chess, and these are not just because of a translation gone bad. It is because the code is really a direct port of the Shogi NNUE training code published at YaneuraOu's GitHub. Also under GPL. And a lot of the terms and references in the comments are actually about Shogi, the game the code was originally written for.

[Dann Corbit]

I would argue that the GPL license is not as much a protection for the software vendors as it is for the software recipients.
GPL programs can freely copy from other GPL programs without any fear of license violation as long as their code remains GPL.
Common decency would include attribution, of course.

On the other hand, it seems the chess community is very concerned about where the ideas come from.
I suppose they might might ban, shun or chastize people who get good ideas from other people.

[Albert Silver]

I would argue that the GPL license is not as much a protection for the software vendors as it is for the software recipients.
GPL programs can freely copy from other GPL programs without any fear of license violation as long as their code remains GPL.
Common decency would include attribution, of course.

On the other hand, it seems the chess community is very concerned about where the ideas come from.
I suppose they might might ban, shun or chastize people who get good ideas from other people.

Standing on the shoulders of midgets

[nabildanial]

https://github.com/AndyGrant/Ethereal/tree/nnue
Pushed NNUE support to a branch. You can load in a file. Stockfish weights are preferred.
If anyone tests this for a rating list, it should be called "Ethereal 12.75 SF-NNUE".
That branch actually works, unlike the Etherlito crap which broke due to memory alignments on POSIX.

Anyone who is willing to share an x64-popcount version?

Here is a compile using the "make release" flag.

[Sylwy]

Already in test:

[cucumber]

I can't imagine that this was an easy thread to make.

You've had a pretty amazing run. I remember cheering Eth's first TCEC runs. I remember how with each release, Eth's progress seemed almost unrelenting. Every time we thought progress was done for, you found some crazy new way to keep progressing.

Since those early days, I think it's safe to say that you've practically singlehandedly added more elo to a chess engine than any single other person. Ethereal has become an incredible feat of engineering, determination, and passion. That's not even getting into the contribution of OpenBench, your research, the voice you've had, the hours of exciting games that Ethereal has provided, etc. Few people have the tenacity to go after a problem like chess for so long, let alone the ability to succeed. So congratulations for actually pulling it off, for pulling it off to the extent that you have, and for helping so many others pull it off to various degrees. Above all, congratulations for reaching the end of this journey.

As saddened as I am about your departure, it's hard not to be excited about what you'll be able to pull off over the next few years. And I'm glad to know that you have one less thing to be stressed about.

Thanks for all you've done, and please keep in touch.