Merge branch 'main' of https://github.com/meta-llama/llama-stack into add_nemo_customizer

llama_stack/apis/agents/agents.py

@@ -36,7 +36,6 @@ from llama_stack.apis.inference import (
 )
 from llama_stack.apis.safety import SafetyViolation
 from llama_stack.apis.tools import ToolDef
-from llama_stack.providers.utils.telemetry.trace_protocol import trace_protocol
 from llama_stack.schema_utils import json_schema_type, register_schema, webmethod
 
 
@@ -383,7 +382,6 @@ class AgentStepResponse(BaseModel):
 
 
 @runtime_checkable
-@trace_protocol
 class Agents(Protocol):
     """Agents API for creating and interacting with agentic systems.
 
@@ -395,7 +393,7 @@ class Agents(Protocol):
     - Agents can also use Memory to retrieve information from knowledge bases. See the RAG Tool and Vector IO APIs for more details.
     """
 
-    @webmethod(route="/agents", method="POST")
+    @webmethod(route="/agents", method="POST", descriptive_name="create_agent")
     async def create_agent(
         self,
         agent_config: AgentConfig,
@@ -407,7 +405,9 @@ class Agents(Protocol):
         """
         ...
 
-    @webmethod(route="/agents/{agent_id}/session/{session_id}/turn", method="POST")
+    @webmethod(
+        route="/agents/{agent_id}/session/{session_id}/turn", method="POST", descriptive_name="create_agent_turn"
+    )
     async def create_agent_turn(
         self,
         agent_id: str,
@@ -439,6 +439,7 @@ class Agents(Protocol):
     @webmethod(
         route="/agents/{agent_id}/session/{session_id}/turn/{turn_id}/resume",
         method="POST",
+        descriptive_name="resume_agent_turn",
     )
     async def resume_agent_turn(
         self,
@@ -501,7 +502,7 @@ class Agents(Protocol):
         """
         ...
 
-    @webmethod(route="/agents/{agent_id}/session", method="POST")
+    @webmethod(route="/agents/{agent_id}/session", method="POST", descriptive_name="create_agent_session")
     async def create_agent_session(
         self,
         agent_id: str,

llama_stack/apis/common/job_types.py

@@ -10,15 +10,15 @@ from pydantic import BaseModel
 from llama_stack.schema_utils import json_schema_type
 
 
-@json_schema_type
-class Job(BaseModel):
-    job_id: str
-
-
-@json_schema_type
 class JobStatus(Enum):
     completed = "completed"
     in_progress = "in_progress"
     failed = "failed"
     scheduled = "scheduled"
     cancelled = "cancelled"
+
+
+@json_schema_type
+class Job(BaseModel):
+    job_id: str
+    status: JobStatus

llama_stack/apis/eval/eval.py

@@ -10,7 +10,7 @@ from pydantic import BaseModel, Field
 from typing_extensions import Annotated
 
 from llama_stack.apis.agents import AgentConfig
-from llama_stack.apis.common.job_types import Job, JobStatus
+from llama_stack.apis.common.job_types import Job
 from llama_stack.apis.inference import SamplingParams, SystemMessage
 from llama_stack.apis.scoring import ScoringResult
 from llama_stack.apis.scoring_functions import ScoringFnParams
@@ -115,7 +115,7 @@ class Eval(Protocol):
         """
 
     @webmethod(route="/eval/benchmarks/{benchmark_id}/jobs/{job_id}", method="GET")
-    async def job_status(self, benchmark_id: str, job_id: str) -> JobStatus:
+    async def job_status(self, benchmark_id: str, job_id: str) -> Job:
         """Get the status of a job.
 
         :param benchmark_id: The ID of the benchmark to run the evaluation on.

llama_stack/apis/post_training/post_training.py

@@ -6,7 +6,7 @@
 
 from datetime import datetime
 from enum import Enum
-from typing import Any, Dict, List, Literal, Optional, Protocol
+from typing import Any, Dict, List, Literal, Optional, Protocol, Union
 
 from pydantic import BaseModel, ConfigDict, Field
 from typing_extensions import Annotated
@@ -89,7 +89,7 @@ class QATFinetuningConfig(BaseModel):
     group_size: int
 
 
-AlgorithmConfig = Annotated[LoraFinetuningConfig | QATFinetuningConfig, Field(discriminator="type")]
+AlgorithmConfig = Annotated[Union[LoraFinetuningConfig, QATFinetuningConfig], Field(discriminator="type")]
 register_schema(AlgorithmConfig, name="AlgorithmConfig")
 
 
@@ -183,7 +183,7 @@ class PostTraining(Protocol):
             description="Model descriptor from `llama model list`",
         ),
         checkpoint_dir: Optional[str] = None,
-        algorithm_config: Optional[LoraFinetuningConfig | QATFinetuningConfig] = None,
+        algorithm_config: Optional[AlgorithmConfig] = None,
     ) -> PostTrainingJob: ...
 
     @webmethod(route="/post-training/preference-optimize", method="POST")

llama_stack/apis/tools/tools.py

@@ -69,7 +69,7 @@ class ToolGroup(Resource):
 
 @json_schema_type
 class ToolInvocationResult(BaseModel):
-    content: InterleavedContent
+    content: Optional[InterleavedContent] = None
     error_message: Optional[str] = None
     error_code: Optional[int] = None
     metadata: Optional[Dict[str, Any]] = None
@@ -140,9 +140,9 @@ class SpecialToolGroup(Enum):
 @runtime_checkable
 @trace_protocol
 class ToolRuntime(Protocol):
-    tool_store: ToolStore
+    tool_store: ToolStore | None = None
 
-    rag_tool: RAGToolRuntime
+    rag_tool: RAGToolRuntime | None = None
 
     # TODO: This needs to be renamed once OPEN API generator name conflict issue is fixed.
     @webmethod(route="/tool-runtime/list-tools", method="GET")

llama_stack/apis/vector_io/vector_io.py

@@ -36,7 +36,7 @@ class VectorDBStore(Protocol):
 @runtime_checkable
 @trace_protocol
 class VectorIO(Protocol):
-    vector_db_store: VectorDBStore
+    vector_db_store: VectorDBStore | None = None
 
     # this will just block now until chunks are inserted, but it should
     # probably return a Job instance which can be polled for completion