futurehouse-client 0.4.5.dev160__tar.gz → 0.4.5.dev201__tar.gz

{futurehouse_client-0.4.5.dev160/src/futurehouse_client.egg-info → futurehouse_client-0.4.5.dev201}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: futurehouse-client
-Version: 0.4.5.dev160
+Version: 0.4.5.dev201
 Summary: A client for interacting with endpoints of the FutureHouse service.
 Author-email: FutureHouse technical staff <hello@futurehouse.org>
 License:                                  Apache License

futurehouse_client-0.4.5.dev201/src/futurehouse_client/models/client.py

@@ -0,0 +1,128 @@
+from typing import Any, Generic, TypeAlias, TypeVar
+
+from aviary.message import Message
+from aviary.tools.base import Tool
+from ldp.agent import Agent
+from ldp.data_structures import Transition
+from ldp.graph.ops import OpResult
+from pydantic import BaseModel, ConfigDict, Field, field_serializer
+
+from .app import Step
+
+T = TypeVar("T")
+
+
+# TODO: revisit this
+# unsure what crow states will return
+# need to revisit after we get more crows deployed
+class BaseState(BaseModel):
+    model_config = ConfigDict(arbitrary_types_allowed=True, extra="allow")
+
+
+class BeforeTransitionState(BaseState):
+    current_state: Any = Field()
+    observations: list[Message] = Field()
+
+
+class InitialState(BaseState):
+    initial_state: Any = Field()
+
+
+class ASVState(BaseState, Generic[T]):
+    action: OpResult[T] = Field()
+    next_state: Any = Field()
+    value: float = Field()
+
+    @field_serializer("action")
+    def serialize_action(self, action: OpResult[T]) -> dict:
+        return action.to_dict()
+
+
+class EnvResetState(BaseState):
+    observations: list[Message] = Field()
+    tools: list[Tool] = Field()
+
+
+class EnvStepState(BaseState):
+    observations: list[Message] = Field()
+    reward: float = Field()
+    done: bool = Field()
+    trunc: bool = Field()
+
+
+class TransitionState(BaseState):
+    transition: Transition = Field()
+
+    @field_serializer("transition")
+    def serialize_transition(self, transition: Transition) -> dict:
+        transition_data = transition.model_dump()
+        return transition_data | {
+            "action": transition.action.to_dict() if transition.action else None,
+        }
+
+
+class GlobalState(BaseState):
+    agent: Agent | None = None
+    env: Any | None = None
+    agent_state: Any | None = None
+    next_agent_state: Any | None = None
+    observations: list = []
+    action: Any | None = None
+    value: float = 0.0
+    last_step_state: Transition | None = None
+
+    def update_observations(self, obs: list[Message]) -> list[Message]:
+        previous_observations = self.observations or []
+        self.observations = obs
+        return previous_observations
+
+    def store_step_state(self, step_state: Transition) -> None:
+        self.last_step_state = step_state
+
+    def update_trajectory_data(self, **kwargs) -> None:
+        for key, value in kwargs.items():
+            setattr(self, key, value)
+
+    def _get_safe_previous_observations(
+        self, current_obs: list[Message] | None = None
+    ) -> list[Message]:
+        if self.last_step_state and self.last_step_state.next_observation:
+            return self.last_step_state.next_observation
+        if self.observations:
+            return self.observations
+        return current_obs or []
+
+    def create_step_state(self, callback_type: Step, **kwargs) -> Transition:
+        defaults = {
+            "timestep": getattr(self.agent, "_timestep", 0) if self.agent else 0,
+            "agent_state": self.agent_state,
+            "next_agent_state": self.next_agent_state or self.agent_state,
+            "observation": self._get_safe_previous_observations(),
+            "next_observation": self.observations or [],
+            "action": self.action,
+            "reward": 0.0,
+            "truncated": False,
+            "done": False,
+            "value": self.value or 0.0,
+            "metadata": {"callback_type": callback_type},
+        }
+
+        for key, value in kwargs.items():
+            if key == "metadata" and isinstance(value, dict):
+                if isinstance(defaults["metadata"], dict):
+                    defaults["metadata"].update(value)
+            else:
+                defaults[key] = value
+
+        return Transition(**defaults)
+
+
+StateType: TypeAlias = (
+    BeforeTransitionState
+    | InitialState
+    | ASVState
+    | EnvResetState
+    | EnvStepState
+    | TransitionState
+    | GlobalState
+)

{futurehouse_client-0.4.5.dev160 → futurehouse_client-0.4.5.dev201}/src/futurehouse_client/version.py

@@ -28,7 +28,7 @@ version_tuple: VERSION_TUPLE
 commit_id: COMMIT_ID
 __commit_id__: COMMIT_ID
 
-__version__ = version = '0.4.5.dev160'
-__version_tuple__ = version_tuple = (0, 4, 5, 'dev160')
+__version__ = version = '0.4.5.dev201'
+__version_tuple__ = version_tuple = (0, 4, 5, 'dev201')
 
-__commit_id__ = commit_id = 'g7b8da3bce'
+__commit_id__ = commit_id = 'g8acb99411'

{futurehouse_client-0.4.5.dev160 → futurehouse_client-0.4.5.dev201/src/futurehouse_client.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: futurehouse-client
-Version: 0.4.5.dev160
+Version: 0.4.5.dev201
 Summary: A client for interacting with endpoints of the FutureHouse service.
 Author-email: FutureHouse technical staff <hello@futurehouse.org>
 License:                                  Apache License

futurehouse_client-0.4.5.dev160/src/futurehouse_client/models/client.py

@@ -1,72 +0,0 @@
-from typing import Any, Generic, TypeAlias, TypeVar
-
-from aviary.message import Message
-from aviary.tools.base import Tool
-from ldp.data_structures import Transition
-from ldp.graph.ops import OpResult
-from pydantic import BaseModel, ConfigDict, Field, field_serializer
-
-T = TypeVar("T")
-
-
-# TODO: revisit this
-# unsure what crow states will return
-# need to revisit after we get more crows deployed
-class BaseState(BaseModel):
-    model_config = ConfigDict(arbitrary_types_allowed=True, extra="allow")
-
-
-class BeforeTransitionState(BaseState):
-    current_state: Any = Field()
-    observations: list[Message] = Field()
-
-
-class InitialState(BaseState):
-    initial_state: Any = Field()
-
-
-class ASVState(BaseState, Generic[T]):
-    action: OpResult[T] = Field()
-    next_state: Any = Field()
-    value: float = Field()
-
-    @field_serializer("action")
-    def serialize_action(self, action: OpResult[T]) -> dict:
-        return action.to_dict()
-
-    @field_serializer("next_state")
-    def serialize_next_state(self, state: Any) -> str:
-        return str(state)
-
-
-class EnvResetState(BaseState):
-    observations: list[Message] = Field()
-    tools: list[Tool] = Field()
-
-
-class EnvStepState(BaseState):
-    observations: list[Message] = Field()
-    reward: float = Field()
-    done: bool = Field()
-    trunc: bool = Field()
-
-
-class TransitionState(BaseState):
-    transition: Transition = Field()
-
-    @field_serializer("transition")
-    def serialize_transition(self, transition: Transition) -> dict:
-        transition_data = transition.model_dump()
-        return transition_data | {
-            "action": transition.action.to_dict() if transition.action else None,
-        }
-
-
-StateType: TypeAlias = (
-    BeforeTransitionState
-    | InitialState
-    | ASVState
-    | EnvResetState
-    | EnvStepState
-    | TransitionState
-)