Fix: model selecton rule in get_model_config_by_type_and_name (#13569)

### What problem does this PR solve?

Fix: model selecton rule in get_model_config_by_type_and_name

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

api/db/joint_services/tenant_model_service.py

@@ -35,11 +35,12 @@ def get_model_config_by_id(tenant_model_id: int) -> dict:
 def get_model_config_by_type_and_name(tenant_id: str, model_type: str, model_name: str):
     if not model_name:
         raise Exception("Model Name is required")
-    model_config = TenantLLMService.get_api_key(tenant_id, model_name)
+    model_type_val = model_type.value if hasattr(model_type, "value") else model_type
+    model_config = TenantLLMService.get_api_key(tenant_id, model_name, model_type_val)
     if not model_config:
         # model_name in format 'name@factory', split model_name and try again
         pure_model_name, fid = TenantLLMService.split_model_name_and_factory(model_name)
-        if model_type == LLMType.EMBEDDING and fid == "Builtin" and "tei-" in os.getenv("COMPOSE_PROFILES", "") and pure_model_name == os.getenv("TEI_MODEL", ""):
+        if model_type_val == LLMType.EMBEDDING.value and fid == "Builtin" and "tei-" in os.getenv("COMPOSE_PROFILES", "") and pure_model_name == os.getenv("TEI_MODEL", ""):
             # configured local embedding model
             embedding_cfg = settings.EMBEDDING_CFG
             config_dict = {
@@ -47,16 +48,22 @@ def get_model_config_by_type_and_name(tenant_id: str, model_type: str, model_nam
                 "api_key": embedding_cfg["api_key"],
                 "llm_name": pure_model_name,
                 "api_base": embedding_cfg["base_url"],
-                "model_type": LLMType.EMBEDDING,
+                "model_type": LLMType.EMBEDDING.value,
             }
         else:
-            model_config = TenantLLMService.get_api_key(tenant_id, pure_model_name)
+            model_config = TenantLLMService.get_api_key(tenant_id, pure_model_name, model_type_val)
             if not model_config:
-                raise LookupError(f"Tenant Model with name {model_name} not found")
+                raise LookupError(f"Tenant Model with name {model_name} and type {model_type_val} not found")
             config_dict = model_config.to_dict()
     else:
         # model_name without @factory
         config_dict = model_config.to_dict()
+    config_model_type = config_dict.get("model_type")
+    config_model_type = config_model_type.value if hasattr(config_model_type, "value") else config_model_type
+    if config_model_type != model_type_val:
+        raise LookupError(
+            f"Tenant Model with name {model_name} has type {config_model_type}, expected {model_type_val}"
+        )
     llm = LLMService.query(llm_name=config_dict["llm_name"])
     if llm:
         config_dict["is_tools"] = llm[0].is_tools

api/db/services/tenant_llm_service.py

@@ -36,12 +36,16 @@ class TenantLLMService(CommonService):
 
     @classmethod
     @DB.connection_context()
-    def get_api_key(cls, tenant_id, model_name):
+    def get_api_key(cls, tenant_id, model_name, model_type=None):
         mdlnm, fid = TenantLLMService.split_model_name_and_factory(model_name)
+        model_type_val = model_type.value if hasattr(model_type, "value") else model_type
+        query_kwargs = {"tenant_id": tenant_id, "llm_name": mdlnm}
+        if model_type_val is not None:
+            query_kwargs["model_type"] = model_type_val
         if not fid:
-            objs = cls.query(tenant_id=tenant_id, llm_name=mdlnm)
+            objs = cls.query(**query_kwargs)
         else:
-            objs = cls.query(tenant_id=tenant_id, llm_name=mdlnm, llm_factory=fid)
+            objs = cls.query(**query_kwargs, llm_factory=fid)
 
         if (not objs) and fid:
             if fid == "LocalAI":
@@ -52,7 +56,8 @@ class TenantLLMService(CommonService):
                 mdlnm += "___OpenAI-API"
             elif fid == "VLLM":
                 mdlnm += "___VLLM"
-            objs = cls.query(tenant_id=tenant_id, llm_name=mdlnm, llm_factory=fid)
+            query_kwargs["llm_name"] = mdlnm
+            objs = cls.query(**query_kwargs, llm_factory=fid)
         if not objs:
             return None
         return objs[0]
@@ -112,10 +117,10 @@ class TenantLLMService(CommonService):
         else:
             assert False, "LLM type error"
 
-        model_config = cls.get_api_key(tenant_id, mdlnm)
+        model_config = cls.get_api_key(tenant_id, mdlnm, llm_type)
         mdlnm, fid = TenantLLMService.split_model_name_and_factory(mdlnm)
         if not model_config:  # for some cases seems fid mismatch
-            model_config = cls.get_api_key(tenant_id, mdlnm)
+            model_config = cls.get_api_key(tenant_id, mdlnm, llm_type)
         if model_config:
             model_config = model_config.to_dict()
         elif llm_type == LLMType.EMBEDDING and fid == "Builtin" and "tei-" in os.getenv("COMPOSE_PROFILES", "") and mdlnm == os.getenv("TEI_MODEL", ""):

test/testcases/test_web_api/test_chunk_app/test_chunk_routes_unit.py

@@ -846,6 +846,9 @@ def test_retrieval_test_branch_matrix_unit(monkeypatch):
             return {"id": "kg-2", "content_with_weight": ""}
 
     monkeypatch.setattr(module, "LLMBundle", lambda *args, **kwargs: llm_calls.append((args, kwargs)) or SimpleNamespace())
+    monkeypatch.setattr(module, "get_model_config_by_type_and_name", lambda *_args, **_kwargs: {"llm_name": "stub-model", "model_type": "chat"})
+    monkeypatch.setattr(module, "get_tenant_default_model_by_type", lambda *_args, **_kwargs: {"llm_name": "stub-model", "model_type": "chat"})
+    monkeypatch.setattr(module, "get_model_config_by_id", lambda *_args, **_kwargs: {"llm_name": "stub-model", "model_type": "embedding"})
     monkeypatch.setattr(module.DocMetadataService, "get_flatted_meta_by_kbs", lambda _kb_ids: [{"meta": "v"}], raising=False)
     monkeypatch.setattr(module, "apply_meta_data_filter", _apply_filter)
     monkeypatch.setattr(module.SearchService, "get_detail", lambda _sid: {"search_config": {"meta_data_filter": {"method": "auto"}, "chat_id": "chat-1"}}, raising=False)

test/testcases/test_web_api/test_conversation_app/test_conversation_routes_unit.py

@@ -180,6 +180,16 @@ async def _read_sse_text(response):
     return "".join(chunks)
 
 
+@pytest.fixture(scope="session")
+def auth():
+    return "unit-auth"
+
+
+@pytest.fixture(scope="session", autouse=True)
+def set_tenant_info():
+    return None
+
+
 @pytest.mark.p2
 def test_set_conversation_update_create_and_errors(monkeypatch):
     module = _load_conversation_module(monkeypatch)
@@ -532,13 +542,13 @@ def test_sequence2txt_validation_and_transcription_paths(monkeypatch):
 
     wav_file = _DummyUploadedFile("audio.wav")
     monkeypatch.setattr(module, "request", _DummyRequest(form={"stream": "false"}, files={"file": wav_file}))
-    monkeypatch.setattr(sys.modules["api.db.joint_services.tenant_model_service"].TenantService, "get_by_id", lambda _uid: (False, None))
+    monkeypatch.setattr(module, "get_tenant_default_model_by_type", lambda *_args, **_kwargs: (_ for _ in ()).throw(LookupError("Tenant not found")))
     res = _run(module.sequence2txt())
     assert res["message"] == "Tenant not found"
 
     wav_file = _DummyUploadedFile("audio.wav")
     monkeypatch.setattr(module, "request", _DummyRequest(form={"stream": "false"}, files={"file": wav_file}))
-    monkeypatch.setattr(sys.modules["api.db.joint_services.tenant_model_service"].TenantService, "get_by_id", lambda _uid: (True, SimpleNamespace(tenant_id="tenant-1", asr_id="")))
+    monkeypatch.setattr(module, "get_tenant_default_model_by_type", lambda *_args, **_kwargs: (_ for _ in ()).throw(Exception("No default speech2text model is set.")))
     res = _run(module.sequence2txt())
     assert res["message"] == "No default speech2text model is set."
 
@@ -551,8 +561,11 @@ def test_sequence2txt_validation_and_transcription_paths(monkeypatch):
 
     wav_file = _DummyUploadedFile("audio.wav")
     monkeypatch.setattr(module, "request", _DummyRequest(form={"stream": "false"}, files={"file": wav_file}))
-    monkeypatch.setattr(sys.modules["api.db.joint_services.tenant_model_service"].TenantService, "get_by_id", lambda _uid: (True, SimpleNamespace(tenant_id="tenant-1", asr_id="asr-model")))
-    monkeypatch.setattr(module.TenantLLMService, "get_api_key", lambda tenant_id, model_name: SimpleNamespace(to_dict=lambda: {"llm_factory": "test", "llm_name": "asr-model"}))
+    monkeypatch.setattr(
+        module,
+        "get_tenant_default_model_by_type",
+        lambda *_args, **_kwargs: {"llm_factory": "test", "llm_name": "asr-model", "model_type": module.LLMType.SPEECH2TEXT.value},
+    )
     monkeypatch.setattr(module, "LLMBundle", lambda *_args, **_kwargs: _SyncAsr())
     monkeypatch.setattr(module.os, "remove", lambda _path: (_ for _ in ()).throw(RuntimeError("remove failed")))
     res = _run(module.sequence2txt())
@@ -593,11 +606,11 @@ def test_sequence2txt_validation_and_transcription_paths(monkeypatch):
 def test_tts_request_parse_entry(monkeypatch):
     module = _load_conversation_module(monkeypatch)
     _set_request_json(monkeypatch, module, {"text": "A。B"})
-    monkeypatch.setattr(sys.modules["api.db.joint_services.tenant_model_service"].TenantService, "get_by_id", lambda _uid: (False, None))
+    monkeypatch.setattr(module, "get_tenant_default_model_by_type", lambda *_args, **_kwargs: (_ for _ in ()).throw(LookupError("Tenant not found")))
     res = _run(module.tts())
     assert res["message"] == "Tenant not found"
 
-    monkeypatch.setattr(sys.modules["api.db.joint_services.tenant_model_service"].TenantService, "get_by_id", lambda _uid: (True, SimpleNamespace(tenant_id="tenant-1", tts_id="")))
+    monkeypatch.setattr(module, "get_tenant_default_model_by_type", lambda *_args, **_kwargs: (_ for _ in ()).throw(Exception("No default tts model is set.")))
     res = _run(module.tts())
     assert res["message"] == "No default tts model is set."
 
@@ -607,8 +620,11 @@ def test_tts_request_parse_entry(monkeypatch):
                 return []
             yield f"chunk-{txt}".encode("utf-8")
 
-    monkeypatch.setattr(sys.modules["api.db.joint_services.tenant_model_service"].TenantService, "get_by_id", lambda _uid: (True, SimpleNamespace(tenant_id="tenant-1", tts_id="tts-x")))
-    monkeypatch.setattr(module.TenantLLMService, "get_api_key", lambda tenant_id, model_name: SimpleNamespace(to_dict=lambda: {"llm_factory": "test", "llm_name": model_name}))
+    monkeypatch.setattr(
+        module,
+        "get_tenant_default_model_by_type",
+        lambda *_args, **_kwargs: {"llm_factory": "test", "llm_name": "tts-x", "model_type": module.LLMType.TTS.value},
+    )
     monkeypatch.setattr(module, "LLMBundle", lambda *_args, **_kwargs: _TTSOk())
     resp = _run(module.tts())
     assert resp.mimetype == "audio/mpeg"
@@ -770,7 +786,11 @@ def test_mindmap_and_related_questions_matrix_unit(monkeypatch):
 
     monkeypatch.setattr(module, "LLMBundle", _fake_bundle)
     monkeypatch.setattr(module, "load_prompt", lambda name: f"prompt-{name}")
-    monkeypatch.setattr(module.TenantLLMService, "get_api_key", lambda tenant_id, model_name: SimpleNamespace(to_dict=lambda: {"llm_factory": "test", "llm_name": model_name}))
+    monkeypatch.setattr(
+        module,
+        "get_model_config_by_type_and_name",
+        lambda *_args, **_kwargs: {"llm_factory": "test", "llm_name": "chat-x", "model_type": module.LLMType.CHAT.value},
+    )
     _set_request_json(monkeypatch, module, {"question": "solar", "search_id": "search-1"})
     res = _run(module.related_questions.__wrapped__())
     assert res["code"] == 0

test/testcases/test_web_api/test_kb_app/test_kb_routes_unit.py

@@ -1064,6 +1064,11 @@ def test_unbind_task_branch_matrix(monkeypatch):
 def test_check_embedding_similarity_threshold_matrix_unit(monkeypatch):
     module = _load_kb_module(monkeypatch)
     route = inspect.unwrap(module.check_embedding)
+    monkeypatch.setattr(
+        module,
+        "get_model_config_by_type_and_name",
+        lambda *_args, **_kwargs: {"llm_factory": "test", "llm_name": "emb-1", "model_type": module.LLMType.EMBEDDING.value},
+    )
     monkeypatch.setattr(module.KnowledgebaseService, "get_by_id", lambda _kb_id: (True, SimpleNamespace(tenant_id="tenant-1")))
     monkeypatch.setattr(module.search, "index_name", lambda _tenant_id: "idx")
 
@@ -1228,6 +1233,11 @@ def test_check_embedding_similarity_threshold_matrix_unit(monkeypatch):
 def test_check_embedding_error_and_empty_sample_paths_unit(monkeypatch):
     module = _load_kb_module(monkeypatch)
     route = inspect.unwrap(module.check_embedding)
+    monkeypatch.setattr(
+        module,
+        "get_model_config_by_type_and_name",
+        lambda *_args, **_kwargs: {"llm_factory": "test", "llm_name": "emb-1", "model_type": module.LLMType.EMBEDDING.value},
+    )
     monkeypatch.setattr(module.KnowledgebaseService, "get_by_id", lambda _kb_id: (True, SimpleNamespace(tenant_id="tenant-1")))
     monkeypatch.setattr(module.search, "index_name", lambda _tenant_id: "idx")
     monkeypatch.setattr(module.random, "sample", lambda population, k: list(population)[:k])