← Older revision

Revision as of 12:47, 18 April 2026

Line 9:

== Pre-trained networks ==

== Pre-trained networks ==

−

HZY initially trained the neural network model on two NVIDIA GeForce RTX 2080 Ti GPUs for about 20 days on 13x13. They then up-trained the network on 19x19 for one day, and on 27x27 for an additional day. [https://zhuanlan-zhihu-com.translate.goog/p/476464087?_x_tr_sl=auto&_x_tr_tl=en According to HZY], the up-trained 19x19 network is relatively reliable, but the 27x27-network is not.

+

HZY initially trained the neural network model (20220618) on two NVIDIA GeForce RTX 2080 Ti GPUs for about 20 days on 13x13. They then up-trained the network on 19x19 for one day, and on 27x27 for an additional day. [https://zhuanlan-zhihu-com.translate.goog/p/476464087?_x_tr_sl=auto&_x_tr_tl=en According to HZY], the up-trained 19x19 network is relatively reliable, but the 27x27-network is not.

−

The newest model, 20240812, is trained on two RTX 4090 GPUs for 2 months on 15x15, 15 days on 19x19, and 3 days on 27x27.

+

A newer model, 20240812, is trained on two RTX 4090 GPUs for 2 months on 15x15, 15 days on 19x19, and 3 days on 27x27. This model uses a b18c384nbt architecture (18 block, 384 channel, [https://github.com/lightvector/KataGo/blob/master/docs/KataGoMethods.md#nested-bottleneck-residual-nets nested bottleneck residual net]).

+

The newest model, 20250131, uses a b28c512nbt architecture. At equal node counts, it is around 100&ndash;200 Elo stronger than 20240812. However, because it is a larger network, each evaluation is around 2.8&times; more costly. (Evaluation time for a network with B blocks and C channels is typically proportional to B &times; C<sup>2</sup>.) Under equal time controls, the two networks are much closer in strength.

Running KataHex requires both a neural net model (the "weights"), and an engine to load the weights. Each neural net is able to play Hex not only at the size it was trained on, but also smaller and larger sizes, as long as the engine is compiled to support the size. (However, the net will not play particularly well on sizes larger than it was trained on.)

Running KataHex requires both a neural net model (the "weights"), and an engine to load the weights. Each neural net is able to play Hex not only at the size it was trained on, but also smaller and larger sizes, as long as the engine is compiled to support the size. (However, the net will not play particularly well on sizes larger than it was trained on.)

Line 25:

Line 27:

'''Program:'''

'''Program:'''

−

* David Wu's original KataGo: <https://github.com/lightvector/KataGo>

+

* David Wu's original KataGo: https://github.com/lightvector/KataGo

−

* HZY's KataHex: <https://github.com/hzyhhzy/KataGo/tree/Hex2024>

+

* HZY's KataHex: https://github.com/hzyhhzy/KataGo/tree/Hex2024

−

* Selinger's modifications: <https://github.com/selinger/katahex/tree/Hex2024>

+

* Selinger's modifications: https://github.com/selinger/katahex/tree/Hex2024

'''Pre-trained network models:'''

'''Pre-trained network models:'''

−

* 20250131 (current strongest net): [https://github.com/hzyhhzy/KataGomo/releases/download/Hex_20250131/hex3_27x_b28.bin.gz hex3_27x_b28.bin.gz]

+

* 20250131 (latest): https://github.com/hzyhhzy/KataGomo/releases/Hex_20250131

−

* 20250131 (supports move limit mode, but slightly weaker): [https://github.com/hzyhhzy/KataGomo/releases/download/Hex_20250131/hex3_mm19x_b28.bin.gz hex3_mm19x_b28.bin.gz]

+

** (weights only): [https://github.com/hzyhhzy/KataGomo/releases/download/Hex_20250131/hex3_27x_b28.bin.gz hex3_27x_b28.bin.gz]

−

* 20240812: [https://drive.usercontent.google.com/download?id=1YeqRvAYs7YjtPh0xBbDnHxeo2xrLEOdX hex27x3.bin.gz]

+

** (supports move limit mode, but slightly weaker): [https://github.com/hzyhhzy/KataGomo/releases/download/Hex_20250131/hex3_mm19x_b28.bin.gz hex3_mm19x_b28.bin.gz]

+

* 20240812: https://github.com/hzyhhzy/KataGomo_fork/releases/Hex_20240812

+

** (weights only): [https://drive.usercontent.google.com/download?id=1YeqRvAYs7YjtPh0xBbDnHxeo2xrLEOdX hex27x3.bin.gz]

* 20220618: [https://drive.google.com/file/d/1xMvP_75xgo0271nQbmlAJ40rvpKiFTgP/view katahex_model_20220618.bin.gz]

* 20220618: [https://drive.google.com/file/d/1xMvP_75xgo0271nQbmlAJ40rvpKiFTgP/view katahex_model_20220618.bin.gz]

'''GUIs:'''

'''GUIs:'''

−

* The 20240812 net is bundled with a modified LizzieYzy GUI: <https://github.com/hzyhhzy/KataGo/releases/tag/Hex_20240812>

+

* The 20240812 net is bundled with a modified LizzieYzy GUI: https://github.com/hzyhhzy/KataGo/releases/tag/Hex_20240812

−

* An older net with the modified LizzieYzy GUI: [https://drive.google.com/file/d/1qbTTmPFiUkM_346DeKS1E9gJR-roNH63/view KataHex_LizzieYZY_20220313.zip].

+

* An older net with the modified LizzieYzy GUI: [https://drive.google.com/file/d/1qbTTmPFiUkM_346DeKS1E9gJR-roNH63/view KataHex_LizzieYZY_20220313.zip]

−

* Selinger's KataHex works with HexGUI: <https://github.com/selinger/hexgui>

+

* Selinger's KataHex works with HexGUI: https://github.com/selinger/hexgui

== See also ==

== See also ==

← Older revision		Revision as of 21:06, 17 April 2026
Line 1:		Line 1:
−	In Hex, as in other strategy games, '''reading''' is the mental process of analyzing a Hex position by considering the possible moves and variations ~~than~~ can occur. By exploring different lines of play and predicting their outcomes, players can determine which moves are most advantageous and make informed decisions. This process often requires backtracking to try all necessary possibilities.	+	In Hex, as in other strategy games, '''reading''' is the mental process of analyzing a Hex position by considering the possible moves and variations that can occur. By exploring different lines of play and predicting their outcomes, players can determine which moves are most advantageous and make informed decisions. This process often requires backtracking to try all necessary possibilities.

	Strong reading skills are essential in Hex, as making a wrong move can quickly lead to a disadvantage or even defeat. One effective way to improve reading skills is to solve [[puzzles]] or play through game scenarios. A good way to do this is with the help of a [[physical Hex sets\|physical Hex board]], or using a Hex editing program that supports variations, such as the [https://hexworld.org/sgf-editor/ HexWorld SGF editor].		Strong reading skills are essential in Hex, as making a wrong move can quickly lead to a disadvantage or even defeat. One effective way to improve reading skills is to solve [[puzzles]] or play through game scenarios. A good way to do this is with the help of a [[physical Hex sets\|physical Hex board]], or using a Hex editing program that supports variations, such as the [https://hexworld.org/sgf-editor/ HexWorld SGF editor].

HexWiki - Recent changes [en]

KataHex

Reading