XiuchengWu
300e86956b
- Add ConsensusSnapshot and ConsensusSlot classes for structured consensus state representation. - Introduce OnlineLanguageLearner for stateful online learning with experience replay and EWC regularization. - Create WorkspaceEvent and ActionEvent classes for serializable event types. - Develop InMemoryVectorMemoryStore for storing and querying workspace events. - Implement ActionPolicy to manage action selection and motor parameterization. - Establish a unified WorkspaceRuntime for managing the agent's lifecycle and interactions. - Enhance tests in test_smoke.py to cover new functionalities and ensure correctness.
26 lines
527 B
Python
26 lines
527 B
Python
"""
|
|
Backward-compatible autonomous runtime exports.
|
|
|
|
The primary implementation now lives in runtime.py so the workspace loop has a
|
|
single home. This module keeps the older import path stable.
|
|
"""
|
|
from .runtime import (
|
|
CuriosityDrive,
|
|
RuntimeStateStore,
|
|
PersistentState,
|
|
SelfModel,
|
|
MetaLearner,
|
|
WorkspaceRuntime,
|
|
AutonomousMind,
|
|
)
|
|
|
|
__all__ = [
|
|
"CuriosityDrive",
|
|
"RuntimeStateStore",
|
|
"PersistentState",
|
|
"SelfModel",
|
|
"MetaLearner",
|
|
"WorkspaceRuntime",
|
|
"AutonomousMind",
|
|
]
|