games.chess.config¶
Chess agent configuration module.
from typing import Any This module provides configuration classes for chess agents, including:
Core game parameters
LLM engine settings
Analysis options
Visualization settings
State schema definition
The configuration system uses Pydantic for validation and default values, making it easy to create and customize chess agent instances.
Classes¶
Configuration class for chess game agents. |
Module Contents¶
- class games.chess.config.ChessConfig¶
Bases:
haive.games.core.config.BaseGameConfig
Configuration class for chess game agents.
This class defines all configuration parameters for a chess agent, including state schema, LLM engines, game settings, and visualization options.
- state_schema¶
The state schema for the game.
- Type:
Type[ChessState]
- engines¶
LLM configurations for players and analyzers.
- Type:
Dict[str, AugLLMConfig]
Examples
>>> # Create a basic configuration >>> config = ChessConfig() >>> >>> # Create a configuration with analysis disabled >>> config = ChessConfig(enable_analysis=False) >>> >>> # Create a configuration with custom LLM engines >>> from haive.core.engine.aug_llm import build_aug_llm >>> engines = { ... "white_player": build_aug_llm("openai", "gpt-4"), ... "black_player": build_aug_llm("anthropic", "claude-3-opus-20240229"), ... } >>> config = ChessConfig(engines=engines)
- class Config¶
Pydantic configuration.
This inner class configures Pydantic behavior for the ChessAgentConfig.
- create_engines_from_player_configs(player_configs)¶
Create engines from player configurations.
- Parameters:
player_configs (dict[str, haive.games.core.agent.player_agent.PlayerAgentConfig])
- Return type:
list[Any]
- create_simple_player_configs()¶
Create player configs from simple model strings.
- Return type:
dict[str, haive.games.core.agent.player_agent.PlayerAgentConfig]
- finalize_config()¶
Finalize configuration after engine setup.
- Return type: