New engine releases & news H1 2023

[Tearth]

Inanis 1.2.0
https://github.com/Tearth/Inanis

[voffka]

Igel 3.3.0 at https://github.com/vshcherbyna/igel/releases/tag/3.3.0 (official executable binaries for Windows and IGN net).

This release brings adequate improvement in strength as a result of a switch to a new network architecture (halfkav2) and re-training network from scratch using existing Igel data as well as newly generated data from Igel 3.2.0.

In order to improve the quality of existing data some work was done to 'sanitize' my old data files. In addition to the sanitization, there were also some improvements in Igel's training data generator such as adding hash table to avoid duplications and more stricts fen checkes that helped to produce a higher quality of Igel 3.2.0 data used for training.

I also experimented with d5 positions that resulted in ability to generate data quickly to get some elo in initial phases of the training. Network is trained of d5 + d8 Igel positions with d16 validation Igel positions.

This version only supports AVX2 builds but I am going to look at adding AVX512 support in future releases.

What's new:

- new network architecture
- new network ign-2-51ba2968

Regression run against Igel 3.2.0 on Fast Time Control (3+0.03)

Code: Select all

Score of Igel 3.3.0 64 BMI2 AVX2 vs Igel 3.2.0 64 BMI2 AVX2: 1713 - 756 - 2531  [0.596] 5000
...      Igel 3.3.0 64 BMI2 AVX2 playing White: 1073 - 258 - 1169  [0.663] 2500
...      Igel 3.3.0 64 BMI2 AVX2 playing Black: 640 - 498 - 1362  [0.528] 2500
...      White vs Black: 1571 - 898 - 2531  [0.567] 5000
Elo difference: 67.3 +/- 6.8, LOS: 100.0 %, DrawRatio: 50.6 %

In case of any questions please ask them in the main thread for Igel at https://talkchess.com/forum3/viewtopic. ... 96#p941596

[Witek]

Caissa 1.5 has been released:
https://github.com/Witek902/Caissa/releases/tag/1.5

Around +50Elo in regular chess and +70Elo in FRC in 8+0.08 time controls. Around +100Elo gain when running on 4 threads in short time controls.

Biggest changes:
* new neural networks (with changed architecture) trained on more positions
* multi-threading optimizations
* neural network evaluation optimizations

[Graham Banks]

Koivisto 9.0

https://github.com/Luecx/Koivisto

[pkrisz]

Hello everyone! I've started programming a chess engine as a hobby back in October and I've originally intended it to be a one afternoon project, but it quickly spiraled out of control
It's not very strong yet, and I wrote a large part of it before noticing the massive amount of resources available for engine development. Thus it has a bunch of questionable design choices, but it still plays much better than I've ever expected and I think it's finally ready for a public release. Should be quite stable.

Details:
- Engine name: Renegade
- Version: 0.7.0
- Architecture: 64-bit (uses popcnt and lzcnt)
- Strength: ~1750
- Author: Krisztián Peőcz
- Country: Hungary
- Website: https://github.com/pkrisz99/Renegade

I welcome any feedback or suggestions you might have! I would also appreciate if it would be tested if there's any free capacity at the moment.

Best regards,
Krisztián

[Gabor Szots]

I welcome any feedback or suggestions you might have! I would also appreciate if it would be tested if there's any free capacity at the moment.

Hi Krisz,

I will try to find that capacity in the near future.

[Dariusz]

Koivisto 9.0
https://github.com/Luecx/Koivisto

Download the x64 versions for Linux and Windows from the project's website: https://github.com/Luecx/Koivisto/releases/tag/v9.0

Mac Apple Silicon & Intel & Linux arm64 and other versions -> feel free download Koivisto 9.0 for from here: https://www.chessengeria.com/post/koivisto-9-0-released

[pkrisz]

I welcome any feedback or suggestions you might have! I would also appreciate if it would be tested if there's any free capacity at the moment.

Hi Krisz,

I will try to find that capacity in the near future.

Great, thank you! I'm excited to see the results.

[Gabor Szots]

Frozenight 6.0.0

https://github.com/MinusKelvin/frozenig ... tag/v6.0.0

Allegedly +300 Elos owing to a bigger network.

[Lana Samson]

Princhess 0.9.0

https://github.com/princesslana/prinche ... /tag/0.9.0

Code: Select all

 Rank Name                          Elo     +/-   Games   Score    Draw
    1 princhess-0.9.0               173      16    2000   73.0%   10.0%
    2 princhess-0.8.0               -14      14    2000   47.9%   11.0%
    3 princhess-0.7.0              -156      16    2000   29.0%    9.8%