PRIVATE ProteusSFX-AI tests

[Eduard Nemeth]

Since anyway they cannot do worse than Ediot, I have decided to test also other Stockfish derivatives, starting from Cool Iris 11.70

We don't need your fraudulent results (without a single PGN). I tested Cool Iris 11.70. In my position test, the engine manages 105 out of 110 solutions. That is outstanding. Eman 9.80 was defeated. Unlike you, I posted all games as PGN. Shove your cheating test up your ass.

[dorsz]

SNEAK PEAK OF THE CURRENT MATCH

Blue Marlin is very good on tests solving and is much faster than ProteusSFX-AI on the same hardware (1.4 Mn/s vs 1.8 Mn/s, 24 vs 27 plies) but probably it's suffering my new Super Books. @dorsz I will give you all the games, so we can investigate what is happening

Yes, Blue Marlin is tactically stong, but now it's far behind top engines in terms of strength.
I think you type @dorsz so I should get a notification, but it does not work on this forum.

[dorsz]

Since anyway they cannot do worse than Ediot, I have decided to test also other Stockfish derivatives, starting from Cool Iris 11.70

Interesting.

[AlexChess]

ProteusSFX-AI 231109 vs Blue Marlin 15.7
Blitz 5+0 Final result: 21-5 / 300 games



@dorsz I'm sending you all the games on Banksiagui. I have added all the won games of both sides to my books

[AlexChess]

BANKSIAGUI 300 BLITZ 5+0 GAMES
231215 UPDATE
2700 games played



Next match against Cool Iris 11.70

[AlexChess]

ProteusSFX-AI 231109 vs Cool Iris 11.70
Blitz 5+0 match Partial results: 3-1 / 98 games



dorsz and Sarona, if you find glitches on the matches that I have sent to you, let us know.

FINALLY THE TRUTH ABOUT OUR MODS REAL STRENGHT using the same minimal hardware for both engines, not silly Playchess and Lichess rankings of engines running on 4 cores versus 88 threads.

After completing the matches against all the known derivatives of Playchess, building two incredible books (for white and black) with all won games of both sides, I will restart my games on Playchess.

But I'd like to talk with skilled developers about AI modules. Actually my Python queries work, but their contribute is minimal due to ChatGPT output not reliable and difficult to exploit applying strategy patterns

[Eduard Nemeth]

A lot of people are currently playing Cool Iris on PlayChess. Nobody plays with ProteusSF. Neither does AlexChess. Because he's afraid. That's why he only plays against himself on the server and cheats on us. He is the biggest cheater on the server. And that's why he has to post nonsensical offline results without PGN. It couldn't be more primitive. Sad.

[IbaiBuR]

If you wanted to make something like that work (the AI modules I mean) , you'd have to pay for the OpenAI API (It is not expensive, I pay less than a euro each month) and fine tune a model that fits your module requirements, that probably would be a way.

I think it is not that difficult but it needs a decent amount of training data.

Regards,

Ibai

PS: By the way, why did you create this 2 new Lichess BOT accounts:
- Alemor64
- No-Stockfish-Bots

[Eduard Nemeth]

BTW, AlexChess' excuse that he only plays with 4 cores on the server, is equally primitive. I sometimes play with less power on the server, with only 1400 kns and a Laptop. And Anton Ces also plays with only 2000 kns and a laptop. Nevertheless, AlexChess is afraid to play against us, and therefore only plays against himself where he cheats.

[Eduard Nemeth]

If you wanted to make something like that work (the AI modules I mean) , you'd have to pay for the OpenAI API (It is not expensive, I pay less than a euro each month) and fine tune a model that fits your module requirements, that probably would be a way.

I think it is not that difficult but it needs a decent amount of training data.

Regards,

Ibai

PS: By the way, why did you create this 2 new Lichess BOT accounts:
- Alemor64
- No-Stockfish-Bots

How do you know who these bots belong to? Another 2 bots from ProteusSF , he already has 8 bots, now 10!