sherpa_ai.memory package

Submodules

sherpa_ai.memory.belief module

class sherpa_ai.memory.belief.Belief

Bases: object

The belief of the agent. it contains

1. events: the events observed by the agent, synchronized with the shared memory

2. internal_events: the internal events generated by the agent through its reasoning process (actions)

property action_description

classmethod from_dict(data)

get(key, default=None)

Get value from the dict, the key can be a dot separated string if the value is nested

get_action(action_name) → BaseAction

get_actions() → List[BaseAction]

get_all_keys()

get_by_type(event_type)

get_context(token_counter: Callable[[str], int])

Get the context of the agent

Parameters:

• token_counter – Token counter

• max_tokens – Maximum number of tokens

Returns:

Context of the agent

Return type:

str

get_dict()

get_histories_excluding_types(exclude_types: list[EventType], token_counter: Callable[[str], int] | None = None, max_tokens=4000)

Get the internal history of the agent without events of excluded_type

Parameters:

• token_counter – Token counter

• max_tokens – Maximum number of tokens

• exclude_types – List of events to be excluded

Returns:

Internal history of the agent with event content separated by newlines. History is truncated if the number of tokens exceeds max_tokens.

Return type:

str

get_internal_history(token_counter: Callable[[str], int])

Get the internal history of the agent

Parameters:

token_counter – Token counter

Returns:

Internal history of the agent with event content separated by newlines. History is truncated if the number of tokens exceeds max_tokens.

Return type:

str

get_state() → str

get_state_obj() → State

has(key)

Check if the key exists in the dict

set(key, value)

Set value in the dict, the key can be a dot separated string if the value is nested

set_actions(actions: List[BaseAction])

set_current_task(task: Event)

update(observation: Event)

update_internal(event_type: EventType, agent: str, content: str)

sherpa_ai.memory.shared_memory module

class sherpa_ai.memory.shared_memory.SharedMemory(objective: str, agent_pool: AgentPool = None)

Bases: object

add(event_type: EventType, agent: str, content: str)

add_event(event: Event)

classmethod from_dict(data, agent_pool)

get_by_type(event_type)

observe(belief: Belief)

Module contents

class sherpa_ai.memory.Belief

Bases: object

The belief of the agent. it contains

1. events: the events observed by the agent, synchronized with the shared memory

2. internal_events: the internal events generated by the agent through its reasoning process (actions)

property action_description

classmethod from_dict(data)

get(key, default=None)

Get value from the dict, the key can be a dot separated string if the value is nested

get_action(action_name) → BaseAction

get_actions() → List[BaseAction]

get_all_keys()

get_by_type(event_type)

get_context(token_counter: Callable[[str], int])

Get the context of the agent

Parameters:

• token_counter – Token counter

• max_tokens – Maximum number of tokens

Returns:

Context of the agent

Return type:

str

get_dict()

get_histories_excluding_types(exclude_types: list[EventType], token_counter: Callable[[str], int] | None = None, max_tokens=4000)

Get the internal history of the agent without events of excluded_type

Parameters:

• token_counter – Token counter

• max_tokens – Maximum number of tokens

• exclude_types – List of events to be excluded

Returns:

Internal history of the agent with event content separated by newlines. History is truncated if the number of tokens exceeds max_tokens.

Return type:

str

get_internal_history(token_counter: Callable[[str], int])

Get the internal history of the agent

Parameters:

token_counter – Token counter

Returns:

Internal history of the agent with event content separated by newlines. History is truncated if the number of tokens exceeds max_tokens.

Return type:

str

get_state() → str

get_state_obj() → State

has(key)

Check if the key exists in the dict

set(key, value)

Set value in the dict, the key can be a dot separated string if the value is nested

set_actions(actions: List[BaseAction])

set_current_task(task: Event)

update(observation: Event)

update_internal(event_type: EventType, agent: str, content: str)

class sherpa_ai.memory.SharedMemory(objective: str, agent_pool: AgentPool = None)

Bases: object

add(event_type: EventType, agent: str, content: str)

add_event(event: Event)

classmethod from_dict(data, agent_pool)

get_by_type(event_type)

observe(belief: Belief)