enhance(client): embed model serving spec into job spec (#3032)

client/starwhale/api/_impl/job/handler.py

@@ -408,7 +408,7 @@ def generate_jobs_yaml(
         yaml_path,
         yaml.safe_dump(
             {
-                name: [h.dict() for h in handlers]
+                name: [h.to_dict() for h in handlers]
                 for name, handlers in expanded_handlers.items()
             },
             default_flow_style=False,

client/starwhale/api/_impl/service/service.py

@@ -5,7 +5,7 @@
 import typing as t
 import functools
 
-from .types.types import ComponentSpec
+from starwhale.base.client.models.models import ApiSpec, ServiceSpec
 
 if sys.version_info >= (3, 9):
     from importlib.resources import files
@@ -27,19 +27,6 @@
 )
 
 
-class ApiSpec(SwBaseModel):
-    uri: str
-    inference_type: str
-    components_hint: t.List[ComponentSpec] = Field(default_factory=list)
-
-
-class ServiceSpec(SwBaseModel):
-    title: t.Optional[str] = None
-    description: t.Optional[str] = None
-    version: str
-    apis: t.List[ApiSpec]
-
-
 class Api(SwBaseModel):
     func: t.Callable
     uri: str
@@ -68,7 +55,7 @@ def to_spec(self) -> ApiSpec | None:
         return ApiSpec(
             uri=self.uri,
             inference_type=self.inference_type.name,
-            components_hint=self.inference_type.components_spec(),
+            components=self.inference_type.components_spec(),
         )
 
 

client/starwhale/api/_impl/service/types/llm.py

@@ -1,11 +1,13 @@
 from __future__ import annotations
 
 import inspect
-from typing import Any, Set, List, Callable, Optional
+from typing import Any, Set, Dict, List, Callable, Optional
 
 from pydantic import BaseModel
 from pydantic.dataclasses import dataclass
 
+from starwhale.base.client.models.models import ComponentSpecValueType
+
 from .types import ServiceType, ComponentSpec
 
 
@@ -28,13 +30,13 @@ class LLMChat(ServiceType):
     name = "llm_chat"
 
     # TODO use pydantic model annotations generated arg_types
-    arg_types = {
-        "user_input": str,
-        "history": list,  # list of Message
-        "top_k": int,
-        "top_p": float,
-        "temperature": float,
-        "max_new_tokens": int,
+    arg_types: Dict[str, ComponentSpecValueType] = {
+        "user_input": ComponentSpecValueType.string,
+        "history": ComponentSpecValueType.list,  # list of Message
+        "top_k": ComponentSpecValueType.int,
+        "top_p": ComponentSpecValueType.float,
+        "temperature": ComponentSpecValueType.float,
+        "max_new_tokens": ComponentSpecValueType.int,
     }
 
     def __init__(self, args: Set | None = None) -> None:
@@ -50,7 +52,7 @@ def __init__(self, args: Set | None = None) -> None:
 
     def components_spec(self) -> List[ComponentSpec]:
         return [
-            ComponentSpec(name=arg, type=self.arg_types[arg].__name__)
+            ComponentSpec(name=arg, component_spec_value_type=self.arg_types[arg])
             for arg in self.args
         ]
 

client/starwhale/api/_impl/service/types/types.py

@@ -3,26 +3,18 @@
 from typing import Any, Dict, List, Callable
 
 from starwhale.utils import console
-from starwhale.base.models.base import SwBaseModel
+from starwhale.base.client.models.models import ComponentSpec, ComponentSpecValueType
 
 Inputs = Any
 Outputs = Any
 
 
-class ComponentSpec(SwBaseModel):
-    name: str
-    type: str
-
-    def __hash__(self) -> int:
-        return hash((self.name, self.type))
-
-
 class ServiceType(abc.ABC):
     """Protocol for service types."""
 
     @property
     @abc.abstractmethod
-    def arg_types(self) -> Dict[str, Any]:
+    def arg_types(self) -> Dict[str, ComponentSpecValueType]:
         ...
 
     @property