feat(vector-io): configurable embedding models for all providers (v2)\n\nAdds embedding_model and embedding_dimension fields to all VectorIOConfig classes.\nRouter respects provider defaults with fallback.\nIntroduces embedding_utils helper.\nComprehensive docs & samples.\nResolves #2729

llama_stack/providers/inline/vector_io/chroma/config.py

@@ -6,12 +6,25 @@
 
 from typing import Any
 
-from pydantic import BaseModel
+from pydantic import BaseModel, Field
 
 
 class ChromaVectorIOConfig(BaseModel):
     db_path: str
+    embedding_model: str | None = Field(
+        default=None,
+        description="Optional default embedding model for this provider. If not specified, will use system default.",
+    )
+    embedding_dimension: int | None = Field(
+        default=None,
+        description="Optional embedding dimension override. Only needed for models with variable dimensions (e.g., Matryoshka embeddings). If not specified, will auto-lookup from model registry.",
+    )
 
     @classmethod
     def sample_run_config(cls, db_path: str = "${env.CHROMADB_PATH}", **kwargs: Any) -> dict[str, Any]:
-        return {"db_path": db_path}
+        return {
+            "db_path": db_path,
+            # Optional: Configure default embedding model for this provider
+            # "embedding_model": "all-MiniLM-L6-v2",
+            # "embedding_dimension": 384,  # Only needed for variable-dimension models
+        }

llama_stack/providers/inline/vector_io/faiss/config.py

@@ -6,7 +6,7 @@
 
 from typing import Any
 
-from pydantic import BaseModel
+from pydantic import BaseModel, Field
 
 from llama_stack.providers.utils.kvstore.config import (
     KVStoreConfig,
@@ -18,6 +18,14 @@ from llama_stack.schema_utils import json_schema_type
 @json_schema_type
 class FaissVectorIOConfig(BaseModel):
     kvstore: KVStoreConfig
+    embedding_model: str | None = Field(
+        default=None,
+        description="Optional default embedding model for this provider. If not specified, will use system default.",
+    )
+    embedding_dimension: int | None = Field(
+        default=None,
+        description="Optional embedding dimension override. Only needed for models with variable dimensions (e.g., Matryoshka embeddings). If not specified, will auto-lookup from model registry.",
+    )
 
     @classmethod
     def sample_run_config(cls, __distro_dir__: str, **kwargs: Any) -> dict[str, Any]:
@@ -25,5 +33,8 @@ class FaissVectorIOConfig(BaseModel):
             "kvstore": SqliteKVStoreConfig.sample_run_config(
                 __distro_dir__=__distro_dir__,
                 db_name="faiss_store.db",
-            )
+            ),
+            # Optional: Configure default embedding model for this provider
+            # "embedding_model": "all-MiniLM-L6-v2",
+            # "embedding_dimension": 384,  # Only needed for variable-dimension models
         }

llama_stack/providers/inline/vector_io/milvus/config.py

@@ -20,6 +20,14 @@ class MilvusVectorIOConfig(BaseModel):
     db_path: str
     kvstore: KVStoreConfig = Field(description="Config for KV store backend (SQLite only for now)")
     consistency_level: str = Field(description="The consistency level of the Milvus server", default="Strong")
+    embedding_model: str | None = Field(
+        default=None,
+        description="Optional default embedding model for this provider. If not specified, will use system default.",
+    )
+    embedding_dimension: int | None = Field(
+        default=None,
+        description="Optional embedding dimension override. Only needed for models with variable dimensions (e.g., Matryoshka embeddings). If not specified, will auto-lookup from model registry.",
+    )
 
     @classmethod
     def sample_run_config(cls, __distro_dir__: str, **kwargs: Any) -> dict[str, Any]:
@@ -29,4 +37,7 @@ class MilvusVectorIOConfig(BaseModel):
                 __distro_dir__=__distro_dir__,
                 db_name="milvus_registry.db",
             ),
+            # Optional: Configure default embedding model for this provider
+            # "embedding_model": "all-MiniLM-L6-v2",
+            # "embedding_dimension": 384,  # Only needed for variable-dimension models
         }

llama_stack/providers/inline/vector_io/sqlite_vec/config.py

@@ -17,6 +17,14 @@ from llama_stack.providers.utils.kvstore.config import (
 class SQLiteVectorIOConfig(BaseModel):
     db_path: str = Field(description="Path to the SQLite database file")
     kvstore: KVStoreConfig = Field(description="Config for KV store backend (SQLite only for now)")
+    embedding_model: str | None = Field(
+        default=None,
+        description="Optional default embedding model for this provider. If not specified, will use system default.",
+    )
+    embedding_dimension: int | None = Field(
+        default=None,
+        description="Optional embedding dimension override. Only needed for models with variable dimensions (e.g., Matryoshka embeddings). If not specified, will auto-lookup from model registry.",
+    )
 
     @classmethod
     def sample_run_config(cls, __distro_dir__: str) -> dict[str, Any]:
@@ -26,4 +34,7 @@ class SQLiteVectorIOConfig(BaseModel):
                 __distro_dir__=__distro_dir__,
                 db_name="sqlite_vec_registry.db",
             ),
+            # Optional: Configure default embedding model for this provider
+            # "embedding_model": "all-MiniLM-L6-v2",
+            # "embedding_dimension": 384,  # Only needed for variable-dimension models
         }