vectorvein 0.2.13__tar.gz → 0.2.15__tar.gz

{vectorvein-0.2.13 → vectorvein-0.2.15}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: vectorvein
-Version: 0.2.13
+Version: 0.2.15
 Summary: VectorVein python SDK
 Author-Email: Anderson <andersonby@163.com>
 License: MIT

{vectorvein-0.2.13 → vectorvein-0.2.15}/pyproject.toml

@@ -18,7 +18,7 @@ description = "VectorVein python SDK"
 name = "vectorvein"
 readme = "README.md"
 requires-python = ">=3.10"
-version = "0.2.13"
+version = "0.2.15"
 
 [project.license]
 text = "MIT"

{vectorvein-0.2.13 → vectorvein-0.2.15}/src/vectorvein/settings/__init__.py

@@ -1,6 +1,7 @@
 # @Author: Bi Ying
 # @Date:   2024-07-27 00:30:56
 import warnings
+from copy import deepcopy
 from typing import List, Optional, Literal
 
 from pydantic import BaseModel, Field
@@ -59,7 +60,7 @@ class Backends(BaseModel):
 
 class Settings(BaseModel):
     VERSION: Optional[str] = Field(
-        default=None, description="Configuration version. If provided, will use the corresponding format."
+        default="2", description="Configuration version. If provided, will use the corresponding format."
     )
     endpoints: List[EndpointSetting] = Field(
         default_factory=list, description="Available endpoints for the LLM service."
@@ -114,8 +115,14 @@ class Settings(BaseModel):
             "xai": defs.XAI_MODELS,
         }
 
+        data = deepcopy(data)
+
         version = data.get("VERSION")
 
+        if len(data) == 0:
+            version = "2"
+            data["backends"] = {}
+
         # If V2 format, model configs are in the backends dictionary
         if version == "2":
             if "backends" not in data:
@@ -124,13 +131,12 @@ class Settings(BaseModel):
             backends = data["backends"]
         else:
             backends = data
-            if len(data) > 0:
-                warnings.warn("You're using vectorvein's deprecated V1 format. Please use V2 format.")
+            warnings.warn("You're using vectorvein's deprecated V1 format. Please use V2 format.")
 
         for model_type, default_models in model_types.items():
             if model_type in backends:
-                model_settings = BackendSettings()
                 user_models = backends[model_type].get("models", {})
+                model_settings = BackendSettings()
                 model_settings.update_models(default_models, user_models)
                 backends[model_type] = model_settings
             else:
@@ -149,6 +155,10 @@ class Settings(BaseModel):
     def load(self, settings_dict: SettingsDict):
         self.__init__(**settings_dict)
 
+    @classmethod
+    def load_from_dict(cls, settings_dict: SettingsDict):
+        return cls(**settings_dict)
+
     def get_endpoint(self, endpoint_id: str) -> EndpointSetting:
         for endpoint in self.endpoints:
             if endpoint.id == endpoint_id:
@@ -165,5 +175,29 @@ class Settings(BaseModel):
         # Compatible with VERSION 1 format
         return getattr(self, backend_name)
 
+    def export(self):
+        if self.VERSION == "2":
+            return super().model_dump(
+                exclude={
+                    "anthropic",
+                    "deepseek",
+                    "gemini",
+                    "groq",
+                    "local",
+                    "minimax",
+                    "mistral",
+                    "moonshot",
+                    "openai",
+                    "qwen",
+                    "yi",
+                    "zhipuai",
+                    "baichuan",
+                    "stepfun",
+                    "xai",
+                }
+            )
+        else:
+            return super().model_dump(exclude={"backends"})
+
 
 settings = Settings()

{vectorvein-0.2.13 → vectorvein-0.2.15}/src/vectorvein/types/__init__.py

@@ -61,8 +61,13 @@ from .settings import (
     BackendSettingsDict,
     EndpointSettingDict,
     SettingsDict,
+    SettingsV1Dict,
+    SettingsV2Dict,
 )
 
+from anthropic.types import ThinkingConfigParam, ThinkingConfigEnabledParam
+from openai.types.chat.completion_create_params import ResponseFormat
+
 
 __all__ = [
     "CONTEXT_LENGTH_CONTROL",
@@ -126,4 +131,9 @@ __all__ = [
     "BackendSettingsDict",
     "EndpointSettingDict",
     "SettingsDict",
+    "SettingsV1Dict",
+    "SettingsV2Dict",
+    "ThinkingConfigParam",
+    "ThinkingConfigEnabledParam",
+    "ResponseFormat",
 ]

{vectorvein-0.2.13 → vectorvein-0.2.15}/src/vectorvein/types/defaults.py

@@ -176,62 +176,6 @@ GROQ_MODELS: Final[Dict[str, Dict[str, Any]]] = {
 # Qwen models
 QWEN_DEFAULT_MODEL: Final[str] = "qwen2.5-72b-instruct"
 QWEN_MODELS: Final[Dict[str, Dict[str, Any]]] = {
-    "qwen1.5-1.8b-chat": {
-        "id": "qwen1.5-1.8b-chat",
-        "context_length": 30000,
-        "max_output_tokens": 4096,
-        "function_call_available": False,
-        "response_format_available": True,
-    },
-    "qwen1.5-4b-chat": {
-        "id": "qwen1.5-4b-chat",
-        "context_length": 30000,
-        "max_output_tokens": 4096,
-        "function_call_available": False,
-        "response_format_available": True,
-    },
-    "qwen1.5-7b-chat": {
-        "id": "qwen1.5-7b-chat",
-        "context_length": 30000,
-        "max_output_tokens": 4096,
-        "function_call_available": False,
-        "response_format_available": True,
-    },
-    "qwen1.5-14b-chat": {
-        "id": "qwen1.5-14b-chat",
-        "context_length": 30000,
-        "max_output_tokens": 4096,
-        "function_call_available": False,
-        "response_format_available": True,
-    },
-    "qwen1.5-32b-chat": {
-        "id": "qwen1.5-32b-chat",
-        "context_length": 30000,
-        "max_output_tokens": 4096,
-        "function_call_available": False,
-        "response_format_available": True,
-    },
-    "qwen1.5-72b-chat": {
-        "id": "qwen1.5-72b-chat",
-        "context_length": 30000,
-        "max_output_tokens": 4096,
-        "function_call_available": False,
-        "response_format_available": True,
-    },
-    "qwen1.5-110b-chat": {
-        "id": "qwen1.5-110b-chat",
-        "context_length": 30000,
-        "max_output_tokens": 4096,
-        "function_call_available": False,
-        "response_format_available": True,
-    },
-    "qwen2-72b-instruct": {
-        "id": "qwen2-72b-instruct",
-        "context_length": 30000,
-        "max_output_tokens": 4096,
-        "function_call_available": False,
-        "response_format_available": True,
-    },
     "qwen2.5-7b-instruct": {
         "id": "qwen2.5-7b-instruct",
         "context_length": 131072,
@@ -282,6 +226,30 @@ QWEN_MODELS: Final[Dict[str, Dict[str, Any]]] = {
         "response_format_available": False,
         "native_multimodal": True,
     },
+    "qwen2.5-vl-72b-instruct": {
+        "id": "qwen2.5-vl-72b-instruct",
+        "context_length": 131072,
+        "max_output_tokens": 8192,
+        "function_call_available": False,
+        "response_format_available": False,
+        "native_multimodal": True,
+    },
+    "qwen2.5-vl-7b-instruct": {
+        "id": "qwen2.5-vl-7b-instruct",
+        "context_length": 131072,
+        "max_output_tokens": 8192,
+        "function_call_available": False,
+        "response_format_available": False,
+        "native_multimodal": True,
+    },
+    "qwen2.5-vl-3b-instruct": {
+        "id": "qwen2.5-vl-3b-instruct",
+        "context_length": 131072,
+        "max_output_tokens": 8192,
+        "function_call_available": False,
+        "response_format_available": False,
+        "native_multimodal": True,
+    },
     "qwen-max": {
         "id": "qwen-max",
         "context_length": 8096,

{vectorvein-0.2.13 → vectorvein-0.2.15}/src/vectorvein/types/settings.py

@@ -101,17 +101,13 @@ class BackendsDict(TypedDict):
     xai: NotRequired[BackendSettingsDict]
 
 
-class SettingsDict(TypedDict):
+class SettingsV1Dict(TypedDict):
     """TypedDict representing the expected structure of the settings dictionary."""
 
-    VERSION: NotRequired[str]
     endpoints: List[EndpointSettingDict]
     token_server: NotRequired[ServerDict]
     rate_limit: NotRequired[RateLimitConfigDict]
 
-    # V2 format: all model backend configs in a single dictionary
-    backends: NotRequired[BackendsDict]
-
     # V1 format: each model backend config
     anthropic: NotRequired[BackendSettingsDict]
     deepseek: NotRequired[BackendSettingsDict]
@@ -128,3 +124,18 @@ class SettingsDict(TypedDict):
     baichuan: NotRequired[BackendSettingsDict]
     stepfun: NotRequired[BackendSettingsDict]
     xai: NotRequired[BackendSettingsDict]
+
+
+class SettingsV2Dict(TypedDict):
+    """TypedDict representing the expected structure of the settings dictionary."""
+
+    VERSION: NotRequired[str]
+    endpoints: List[EndpointSettingDict]
+    token_server: NotRequired[ServerDict]
+    rate_limit: NotRequired[RateLimitConfigDict]
+
+    # V2 format: all model backend configs in a single dictionary
+    backends: NotRequired[BackendsDict]
+
+
+SettingsDict = Union[SettingsV1Dict, SettingsV2Dict]