resource oriented object design for models

llama_stack/distribution/datatypes.py

@@ -31,7 +31,7 @@ RoutingKey = Union[str, List[str]]
 
 
 RoutableObject = Union[
-    ModelDef,
+    Model,
     Shield,
     MemoryBankDef,
     DatasetDef,
@@ -41,7 +41,7 @@ RoutableObject = Union[
 
 RoutableObjectWithProvider = Annotated[
     Union[
-        ModelDefWithProvider,
+        Model,
         Shield,
         MemoryBankDefWithProvider,
         DatasetDefWithProvider,

llama_stack/distribution/routers/routers.py

@@ -71,7 +71,7 @@ class InferenceRouter(Inference):
     async def shutdown(self) -> None:
         pass
 
-    async def register_model(self, model: ModelDef) -> None:
+    async def register_model(self, model: Model) -> None:
         await self.routing_table.register_model(model)
 
     async def chat_completion(

llama_stack/distribution/routers/routing_tables.py

@@ -84,8 +84,6 @@ class CommonRoutingTableImpl(RoutingTable):
             api = get_impl_api(p)
             if api == Api.inference:
                 p.model_store = self
-                models = await p.list_models()
-                await add_objects(models, pid, ModelDefWithProvider)
             elif api == Api.safety:
                 p.shield_store = self
 
@@ -198,13 +196,13 @@ class CommonRoutingTableImpl(RoutingTable):
 
 
 class ModelsRoutingTable(CommonRoutingTableImpl, Models):
-    async def list_models(self) -> List[ModelDefWithProvider]:
+    async def list_models(self) -> List[Model]:
         return await self.get_all_with_type("model")
 
-    async def get_model(self, identifier: str) -> Optional[ModelDefWithProvider]:
+    async def get_model(self, identifier: str) -> Optional[Model]:
         return await self.get_object_by_identifier(identifier)
 
-    async def register_model(self, model: ModelDefWithProvider) -> None:
+    async def register_model(self, model: Model) -> None:
         await self.register_object(model)
 
 

llama_stack/distribution/store/tests/test_registry.py

@@ -9,7 +9,7 @@ import os
 import pytest
 import pytest_asyncio
 from llama_stack.distribution.store import *  # noqa F403
-from llama_stack.apis.inference import ModelDefWithProvider
+from llama_stack.apis.inference import Model
 from llama_stack.apis.memory_banks import VectorMemoryBankDef
 from llama_stack.providers.utils.kvstore import kvstore_impl, SqliteKVStoreConfig
 from llama_stack.distribution.datatypes import *  # noqa F403
@@ -50,9 +50,8 @@ def sample_bank():
 
 @pytest.fixture
 def sample_model():
-    return ModelDefWithProvider(
+    return Model(
         identifier="test_model",
-        llama_model="Llama3.2-3B-Instruct",
         provider_id="test-provider",
     )
 
@@ -84,7 +83,6 @@ async def test_basic_registration(registry, sample_bank, sample_model):
     assert len(results) == 1
     result_model = results[0]
     assert result_model.identifier == sample_model.identifier
-    assert result_model.llama_model == sample_model.llama_model
     assert result_model.provider_id == sample_model.provider_id