Merge remote-tracking branch 'origin/main' into dependabot/uv/openai-2.5.0

llama_stack/apis/agents/openai_responses.py

@@ -545,6 +545,7 @@ class OpenAIResponseObject(BaseModel):
     :param tools: (Optional) An array of tools the model may call while generating a response.
     :param truncation: (Optional) Truncation strategy applied to the response
     :param usage: (Optional) Token usage information for the response
+    :param instructions: (Optional) System message inserted into the model's context
     """
 
     created_at: int
@@ -564,6 +565,7 @@ class OpenAIResponseObject(BaseModel):
     tools: list[OpenAIResponseTool] | None = None
     truncation: str | None = None
     usage: OpenAIResponseUsage | None = None
+    instructions: str | None = None
 
 
 @json_schema_type

llama_stack/apis/datatypes.py

@@ -121,6 +121,7 @@ class Api(Enum, metaclass=DynamicApiMeta):
 
     models = "models"
     shields = "shields"
+    vector_stores = "vector_stores"  # only used for routing table
     datasets = "datasets"
     scoring_functions = "scoring_functions"
     benchmarks = "benchmarks"

llama_stack/apis/inference/inference.py

@@ -1234,9 +1234,10 @@ class Inference(InferenceProvider):
 
     Llama Stack Inference API for generating completions, chat completions, and embeddings.
 
-    This API provides the raw interface to the underlying models. Two kinds of models are supported:
+    This API provides the raw interface to the underlying models. Three kinds of models are supported:
     - LLM models: these models generate "raw" and "chat" (conversational) completions.
     - Embedding models: these models generate embeddings to be used for semantic search.
+    - Rerank models: these models reorder the documents based on their relevance to a query.
     """
 
     @webmethod(route="/openai/v1/chat/completions", method="GET", level=LLAMA_STACK_API_V1, deprecated=True)

llama_stack/apis/models/models.py

@@ -27,10 +27,12 @@ class ModelType(StrEnum):
     """Enumeration of supported model types in Llama Stack.
     :cvar llm: Large language model for text generation and completion
     :cvar embedding: Embedding model for converting text to vector representations
+    :cvar rerank: Reranking model for reordering documents based on their relevance to a query
     """
 
     llm = "llm"
     embedding = "embedding"
+    rerank = "rerank"
 
 
 @json_schema_type

llama_stack/apis/resource.py

@@ -13,7 +13,7 @@ from pydantic import BaseModel, Field
 class ResourceType(StrEnum):
     model = "model"
     shield = "shield"
-    vector_db = "vector_db"
+    vector_store = "vector_store"
     dataset = "dataset"
     scoring_function = "scoring_function"
     benchmark = "benchmark"
@@ -34,4 +34,4 @@ class Resource(BaseModel):
 
     provider_id: str = Field(description="ID of the provider that owns this resource")
 
-    type: ResourceType = Field(description="Type of resource (e.g. 'model', 'shield', 'vector_db', etc.)")
+    type: ResourceType = Field(description="Type of resource (e.g. 'model', 'shield', 'vector_store', etc.)")

llama_stack/apis/vector_io/vector_io.py

@@ -15,7 +15,7 @@ from fastapi import Body
 from pydantic import BaseModel, Field
 
 from llama_stack.apis.inference import InterleavedContent
-from llama_stack.apis.vector_dbs import VectorDB
+from llama_stack.apis.vector_stores import VectorStore
 from llama_stack.apis.version import LLAMA_STACK_API_V1
 from llama_stack.providers.utils.telemetry.trace_protocol import trace_protocol
 from llama_stack.providers.utils.vector_io.vector_utils import generate_chunk_id
@@ -140,6 +140,7 @@ class VectorStoreFileCounts(BaseModel):
     total: int
 
 
+# TODO: rename this as OpenAIVectorStore
 @json_schema_type
 class VectorStoreObject(BaseModel):
     """OpenAI Vector Store object.
@@ -517,17 +518,18 @@ class OpenAICreateVectorStoreFileBatchRequestWithExtraBody(BaseModel, extra="all
     chunking_strategy: VectorStoreChunkingStrategy | None = None
 
 
-class VectorDBStore(Protocol):
-    def get_vector_db(self, vector_db_id: str) -> VectorDB | None: ...
+class VectorStoreTable(Protocol):
+    def get_vector_store(self, vector_store_id: str) -> VectorStore | None: ...
 
 
 @runtime_checkable
 @trace_protocol
 class VectorIO(Protocol):
-    vector_db_store: VectorDBStore | None = None
+    vector_store_table: VectorStoreTable | 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
+    # TODO: rename vector_db_id to vector_store_id once Stainless is working
     @webmethod(route="/vector-io/insert", method="POST", level=LLAMA_STACK_API_V1)
     async def insert_chunks(
         self,
@@ -546,6 +548,7 @@ class VectorIO(Protocol):
         """
         ...
 
+    # TODO: rename vector_db_id to vector_store_id once Stainless is working
     @webmethod(route="/vector-io/query", method="POST", level=LLAMA_STACK_API_V1)
     async def query_chunks(
         self,

llama_stack/apis/vector_stores/__init__.py

@@ -4,4 +4,4 @@
 # This source code is licensed under the terms described in the LICENSE file in
 # the root directory of this source tree.
 
-from .vector_dbs import *
+from .vector_stores import *

llama_stack/apis/vector_stores/vector_stores.py

@@ -9,53 +9,43 @@ from typing import Literal
 from pydantic import BaseModel
 
 from llama_stack.apis.resource import Resource, ResourceType
-from llama_stack.schema_utils import json_schema_type
 
 
-@json_schema_type
-class VectorDB(Resource):
+# Internal resource type for storing the vector store routing and other information
+class VectorStore(Resource):
     """Vector database resource for storing and querying vector embeddings.
 
-    :param type: Type of resource, always 'vector_db' for vector databases
+    :param type: Type of resource, always 'vector_store' for vector stores
     :param embedding_model: Name of the embedding model to use for vector generation
     :param embedding_dimension: Dimension of the embedding vectors
     """
 
-    type: Literal[ResourceType.vector_db] = ResourceType.vector_db
+    type: Literal[ResourceType.vector_store] = ResourceType.vector_store
 
     embedding_model: str
     embedding_dimension: int
-    vector_db_name: str | None = None
+    vector_store_name: str | None = None
 
     @property
-    def vector_db_id(self) -> str:
+    def vector_store_id(self) -> str:
         return self.identifier
 
     @property
-    def provider_vector_db_id(self) -> str | None:
+    def provider_vector_store_id(self) -> str | None:
         return self.provider_resource_id
 
 
-class VectorDBInput(BaseModel):
+class VectorStoreInput(BaseModel):
     """Input parameters for creating or configuring a vector database.
 
-    :param vector_db_id: Unique identifier for the vector database
+    :param vector_store_id: Unique identifier for the vector store
     :param embedding_model: Name of the embedding model to use for vector generation
     :param embedding_dimension: Dimension of the embedding vectors
-    :param provider_vector_db_id: (Optional) Provider-specific identifier for the vector database
+    :param provider_vector_store_id: (Optional) Provider-specific identifier for the vector store
     """
 
-    vector_db_id: str
+    vector_store_id: str
     embedding_model: str
     embedding_dimension: int
     provider_id: str | None = None
-    provider_vector_db_id: str | None = None
-
-
-class ListVectorDBsResponse(BaseModel):
-    """Response from listing vector databases.
-
-    :param data: List of vector databases
-    """
-
-    data: list[VectorDB]
+    provider_vector_store_id: str | None = None