haive.games.tic_tac_toe.configurable_engines

Configurable Tic Tac Toe engines using the new player agent system.

This module provides Tic Tac Toe engine configurations that use configurable player agents instead of hardcoded LLM configurations.

Functions

create_configurable_tic_tac_toe_engines(player_configs)	Create Tic Tac Toe engines from configurable player agents.
create_simple_tic_tac_toe_player_configs([x_model, ...])	Create simple player configurations for Tic Tac Toe.
create_tic_tac_toe_analysis_prompt(player_symbol)	Create a Tic Tac Toe analysis prompt for the specified player.
create_tic_tac_toe_engines_from_models([x_model, ...])	Create Tic Tac Toe engines using simple model strings.
create_tic_tac_toe_move_prompt(player_symbol)	Create a Tic Tac Toe move prompt for the specified player.
get_example_tic_tac_toe_engines(config_name)	Get example Tic Tac Toe engine configuration by name.
get_tic_tac_toe_role_definitions()	Get role definitions for Tic Tac Toe players and analyzers.

Module Contents

haive.games.tic_tac_toe.configurable_engines.create_configurable_tic_tac_toe_engines(player_configs)

Create Tic Tac Toe engines from configurable player agents.

Parameters:

player_configs (dict[str, haive.games.core.agent.player_agent.PlayerAgentConfig]) – Dictionary of role name to player configuration

Returns:

Dictionary of configured engines

Return type:

Dict[str, AugLLMConfig]

Example

>>> configs = {
...     "X_player": PlayerAgentConfig(llm_config="gpt-4"),
...     "O_player": PlayerAgentConfig(llm_config="claude-3-opus"),
...     "X_analyzer": PlayerAgentConfig(llm_config="gpt-4"),
...     "O_analyzer": PlayerAgentConfig(llm_config="claude-3-opus"),
... }
>>> engines = create_configurable_tic_tac_toe_engines(configs)

haive.games.tic_tac_toe.configurable_engines.create_simple_tic_tac_toe_player_configs(x_model='gpt-4o', o_model='claude-3-5-sonnet-20240620', temperature=0.3)

Create simple player configurations for Tic Tac Toe.

Parameters:

• x_model (str) – Model for X player

• o_model (str) – Model for O player

• temperature (float) – Temperature for both players

Returns:

Player configurations

Return type:

Dict[str, PlayerAgentConfig]

haive.games.tic_tac_toe.configurable_engines.create_tic_tac_toe_analysis_prompt(player_symbol)

Create a Tic Tac Toe analysis prompt for the specified player.

Parameters:

player_symbol (str) – Player symbol (“X” or “O”)

Returns:

Prompt template for position analysis

Return type:

ChatPromptTemplate

haive.games.tic_tac_toe.configurable_engines.create_tic_tac_toe_engines_from_models(x_model='gpt-4o', o_model='claude-3-5-sonnet-20240620', temperature=0.3)

Create Tic Tac Toe engines using simple model strings.

Parameters:

• x_model (str) – Model for X player and analyzer

• o_model (str) – Model for O player and analyzer

• temperature (float) – Temperature for all engines

Returns:

Dictionary of engines

Return type:

Dict[str, AugLLMConfig]

haive.games.tic_tac_toe.configurable_engines.create_tic_tac_toe_move_prompt(player_symbol)

Create a Tic Tac Toe move prompt for the specified player.

Parameters:

player_symbol (str) – Player symbol (“X” or “O”)

Returns:

Prompt template for move generation

Return type:

ChatPromptTemplate

haive.games.tic_tac_toe.configurable_engines.get_example_tic_tac_toe_engines(config_name)

Get example Tic Tac Toe engine configuration by name.

Parameters:

config_name (str) – Name of the configuration from EXAMPLE_TTT_CONFIGS

Returns:

Dictionary of engines

Return type:

Dict[str, AugLLMConfig]

Available configs: gpt_vs_claude, gpt_only, claude_only, budget, mixed

haive.games.tic_tac_toe.configurable_engines.get_tic_tac_toe_role_definitions()

Get role definitions for Tic Tac Toe players and analyzers.

Returns:

Dictionary of role definitions

Return type:

Dict[str, GamePlayerRole]