vectorvein 0.2.11__py3-none-any.whl → 0.2.13__py3-none-any.whl

vectorvein/chat_clients/openai_compatible_client.py

@@ -99,7 +99,7 @@ class OpenAICompatibleChatClient(BaseChatClient):
             return AzureOpenAI(
                 azure_endpoint=self.endpoint.api_base,
                 api_key=self.endpoint.api_key,
-                api_version="2024-12-01-preview",
+                api_version="2025-01-01-preview",
                 http_client=self.http_client,
             )
         else:
@@ -574,7 +574,7 @@ class AsyncOpenAICompatibleChatClient(BaseAsyncChatClient):
             return AsyncAzureOpenAI(
                 azure_endpoint=self.endpoint.api_base,
                 api_key=self.endpoint.api_key,
-                api_version="2024-12-01-preview",
+                api_version="2025-01-01-preview",
                 http_client=self.http_client,
             )
         else:

vectorvein/settings/__init__.py

@@ -1,11 +1,13 @@
 # @Author: Bi Ying
 # @Date:   2024-07-27 00:30:56
-from typing import List, Dict, Optional, Literal
+import warnings
+from typing import List, Optional, Literal
 
 from pydantic import BaseModel, Field
 
 from ..types import defaults as defs
 from ..types.enums import BackendType
+from ..types.settings import SettingsDict
 from ..types.llm_parameters import BackendSettings, EndpointSetting
 
 
@@ -35,12 +37,8 @@ class Server(BaseModel):
     url: Optional[str]
 
 
-class Settings(BaseModel):
-    endpoints: List[EndpointSetting] = Field(
-        default_factory=list, description="Available endpoints for the LLM service."
-    )
-    token_server: Optional[Server] = Field(default=None, description="Token server address. Format: host:port")
-    rate_limit: Optional[RateLimitConfig] = Field(default=None, description="Rate limit settings.")
+class Backends(BaseModel):
+    """Model containing all backend configurations in one place."""
 
     anthropic: BackendSettings = Field(default_factory=BackendSettings, description="Anthropic models settings.")
     deepseek: BackendSettings = Field(default_factory=BackendSettings, description="Deepseek models settings.")
@@ -58,6 +56,45 @@ class Settings(BaseModel):
     stepfun: BackendSettings = Field(default_factory=BackendSettings, description="StepFun models settings.")
     xai: BackendSettings = Field(default_factory=BackendSettings, description="XAI models settings.")
 
+
+class Settings(BaseModel):
+    VERSION: Optional[str] = Field(
+        default=None, description="Configuration version. If provided, will use the corresponding format."
+    )
+    endpoints: List[EndpointSetting] = Field(
+        default_factory=list, description="Available endpoints for the LLM service."
+    )
+    token_server: Optional[Server] = Field(default=None, description="Token server address. Format: host:port")
+    rate_limit: Optional[RateLimitConfig] = Field(default=None, description="Rate limit settings.")
+
+    # V2 format: all model backend configs in a single dictionary
+    backends: Optional[Backends] = Field(default=None, description="All model backends in one place (V2 format).")
+
+    # V1 format: each model backend config
+    anthropic: Optional[BackendSettings] = Field(
+        default_factory=BackendSettings, description="Anthropic models settings."
+    )
+    deepseek: Optional[BackendSettings] = Field(
+        default_factory=BackendSettings, description="Deepseek models settings."
+    )
+    gemini: Optional[BackendSettings] = Field(default_factory=BackendSettings, description="Gemini models settings.")
+    groq: Optional[BackendSettings] = Field(default_factory=BackendSettings, description="Groq models settings.")
+    local: Optional[BackendSettings] = Field(default_factory=BackendSettings, description="Local models settings.")
+    minimax: Optional[BackendSettings] = Field(default_factory=BackendSettings, description="Minimax models settings.")
+    mistral: Optional[BackendSettings] = Field(default_factory=BackendSettings, description="Mistral models settings.")
+    moonshot: Optional[BackendSettings] = Field(
+        default_factory=BackendSettings, description="Moonshot models settings."
+    )
+    openai: Optional[BackendSettings] = Field(default_factory=BackendSettings, description="OpenAI models settings.")
+    qwen: Optional[BackendSettings] = Field(default_factory=BackendSettings, description="Qwen models settings.")
+    yi: Optional[BackendSettings] = Field(default_factory=BackendSettings, description="Yi models settings.")
+    zhipuai: Optional[BackendSettings] = Field(default_factory=BackendSettings, description="Zhipuai models settings.")
+    baichuan: Optional[BackendSettings] = Field(
+        default_factory=BackendSettings, description="Baichuan models settings."
+    )
+    stepfun: Optional[BackendSettings] = Field(default_factory=BackendSettings, description="StepFun models settings.")
+    xai: Optional[BackendSettings] = Field(default_factory=BackendSettings, description="XAI models settings.")
+
     def __init__(self, **data):
         model_types = {
             "anthropic": defs.ANTHROPIC_MODELS,
@@ -74,15 +111,30 @@ class Settings(BaseModel):
             "zhipuai": defs.ZHIPUAI_MODELS,
             "baichuan": defs.BAICHUAN_MODELS,
             "stepfun": defs.STEPFUN_MODELS,
+            "xai": defs.XAI_MODELS,
         }
 
+        version = data.get("VERSION")
+
+        # If V2 format, model configs are in the backends dictionary
+        if version == "2":
+            if "backends" not in data:
+                raise ValueError("backends is required in V2 format.")
+
+            backends = data["backends"]
+        else:
+            backends = data
+            if len(data) > 0:
+                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 data:
+            if model_type in backends:
                 model_settings = BackendSettings()
-                model_settings.update_models(default_models, data[model_type].get("models", {}))
-                data[model_type] = model_settings
+                user_models = backends[model_type].get("models", {})
+                model_settings.update_models(default_models, user_models)
+                backends[model_type] = model_settings
             else:
-                data[model_type] = BackendSettings(models=default_models)
+                backends[model_type] = BackendSettings(models=default_models)
 
         for endpoint in data.get("endpoints", []):
             if not endpoint.get("api_base"):
@@ -94,7 +146,7 @@ class Settings(BaseModel):
 
         super().__init__(**data)
 
-    def load(self, settings_dict: Dict):
+    def load(self, settings_dict: SettingsDict):
         self.__init__(**settings_dict)
 
     def get_endpoint(self, endpoint_id: str) -> EndpointSetting:
@@ -104,7 +156,14 @@ class Settings(BaseModel):
         raise ValueError(f"Endpoint {endpoint_id} not found.")
 
     def get_backend(self, backend: BackendType) -> BackendSettings:
-        return getattr(self, backend.value.lower())
+        backend_name = backend.value.lower()
+
+        # Use VERSION 2 format backends field first
+        if self.VERSION == "2" and self.backends is not None:
+            return getattr(self.backends, backend_name)
+
+        # Compatible with VERSION 1 format
+        return getattr(self, backend_name)
 
 
 settings = Settings()

vectorvein/types/llm_parameters.py

@@ -2,6 +2,7 @@
 # @Date:   2024-07-26 23:48:04
 from typing import List, Dict, Optional, Union, Iterable
 
+import httpx
 from pydantic import BaseModel, Field
 
 from anthropic._types import NotGiven as AnthropicNotGiven
@@ -10,6 +11,7 @@ from anthropic.types import ToolParam as AnthropicToolParam
 from anthropic.types import ThinkingConfigParam, ThinkingConfigEnabledParam
 from anthropic.types.message_create_params import ToolChoice as AnthropicToolChoice
 
+from openai import OpenAI, AsyncOpenAI, AzureOpenAI, AsyncAzureOpenAI
 from openai._types import NotGiven as OpenAINotGiven
 from openai._types import NOT_GIVEN as OPENAI_NOT_GIVEN
 from openai.types.chat.completion_create_params import ResponseFormat
@@ -45,6 +47,48 @@ class EndpointSetting(BaseModel):
     )
     proxy: Optional[str] = Field(None, description="The proxy URL for the endpoint.")
 
+    def model_list(self):
+        http_client = httpx.Client(proxy=self.proxy) if self.proxy is not None else None
+
+        if self.is_azure:
+            if self.api_base is None:
+                raise ValueError("Azure endpoint is not set")
+            _client = AzureOpenAI(
+                azure_endpoint=self.api_base,
+                api_key=self.api_key,
+                api_version="2025-01-01-preview",
+                http_client=http_client,
+            )
+        else:
+            _client = OpenAI(
+                api_key=self.api_key,
+                base_url=self.api_base,
+                http_client=http_client,
+            )
+
+        return _client.models.list().model_dump()
+
+    async def amodel_list(self):
+        http_client = httpx.AsyncClient(proxy=self.proxy) if self.proxy is not None else None
+
+        if self.is_azure:
+            if self.api_base is None:
+                raise ValueError("Azure endpoint is not set")
+            _client = AsyncAzureOpenAI(
+                azure_endpoint=self.api_base,
+                api_key=self.api_key,
+                api_version="2025-01-01-preview",
+                http_client=http_client,
+            )
+        else:
+            _client = AsyncOpenAI(
+                api_key=self.api_key,
+                base_url=self.api_base,
+                http_client=http_client,
+            )
+
+        return (await _client.models.list()).model_dump()
+
 
 class ModelSetting(BaseModel):
     id: str = Field(..., description="The id of the model.")

vectorvein/types/settings.py

@@ -21,8 +21,8 @@ class RateLimitConfigDict(TypedDict):
 
     enabled: bool
     backend: Literal["memory", "redis", "diskcache"]
-    redis: Optional[RedisConfigDict]
-    diskcache: Optional[DiskCacheConfigDict]
+    redis: NotRequired[RedisConfigDict]
+    diskcache: NotRequired[DiskCacheConfigDict]
     default_rpm: int
     default_tpm: int
 
@@ -32,7 +32,7 @@ class ServerDict(TypedDict):
 
     host: str
     port: int
-    url: Optional[str]
+    url: NotRequired[str]
 
 
 class EndpointOptionDict(TypedDict):
@@ -50,11 +50,11 @@ class ModelConfigDict(TypedDict):
 
     id: str
     endpoints: List[Union[str, EndpointOptionDict]]
-    function_call_available: bool
-    response_format_available: bool
-    native_multimodal: bool
-    context_length: int
-    max_output_tokens: Optional[int]
+    function_call_available: NotRequired[bool]
+    response_format_available: NotRequired[bool]
+    native_multimodal: NotRequired[bool]
+    context_length: NotRequired[int]
+    max_output_tokens: NotRequired[int]
 
 
 class BackendSettingsDict(TypedDict):
@@ -67,39 +67,64 @@ class EndpointSettingDict(TypedDict):
     """TypedDict representing the EndpointSetting structure."""
 
     id: str
-    api_base: Optional[str]
-    api_key: str
-    region: Optional[str]
-    api_schema_type: Optional[str]
-    credentials: Optional[dict]
-    is_azure: Optional[bool]
-    is_vertex: Optional[bool]
-    is_bedrock: Optional[bool]
-    rpm: Optional[int]
-    tpm: Optional[int]
-    concurrent_requests: Optional[int]
-    proxy: Optional[str]
+    api_base: NotRequired[Optional[str]]
+    api_key: NotRequired[str]
+    region: NotRequired[str]
+    api_schema_type: NotRequired[str]
+    credentials: NotRequired[dict]
+    is_azure: NotRequired[bool]
+    is_vertex: NotRequired[bool]
+    is_bedrock: NotRequired[bool]
+    rpm: NotRequired[int]
+    tpm: NotRequired[int]
+    concurrent_requests: NotRequired[int]
+    proxy: NotRequired[str]
+
+
+class BackendsDict(TypedDict):
+    """TypedDict representing all model backends in a single dictionary."""
+
+    anthropic: NotRequired[BackendSettingsDict]
+    deepseek: NotRequired[BackendSettingsDict]
+    gemini: NotRequired[BackendSettingsDict]
+    groq: NotRequired[BackendSettingsDict]
+    local: NotRequired[BackendSettingsDict]
+    minimax: NotRequired[BackendSettingsDict]
+    mistral: NotRequired[BackendSettingsDict]
+    moonshot: NotRequired[BackendSettingsDict]
+    openai: NotRequired[BackendSettingsDict]
+    qwen: NotRequired[BackendSettingsDict]
+    yi: NotRequired[BackendSettingsDict]
+    zhipuai: NotRequired[BackendSettingsDict]
+    baichuan: NotRequired[BackendSettingsDict]
+    stepfun: NotRequired[BackendSettingsDict]
+    xai: NotRequired[BackendSettingsDict]
 
 
 class SettingsDict(TypedDict):
     """TypedDict representing the expected structure of the settings dictionary."""
 
+    VERSION: NotRequired[str]
     endpoints: List[EndpointSettingDict]
-    token_server: Optional[ServerDict]
-    rate_limit: Optional[RateLimitConfigDict]
-    # 各模型后端配置
-    anthropic: BackendSettingsDict
-    deepseek: BackendSettingsDict
-    gemini: BackendSettingsDict
-    groq: BackendSettingsDict
-    local: BackendSettingsDict
-    minimax: BackendSettingsDict
-    mistral: BackendSettingsDict
-    moonshot: BackendSettingsDict
-    openai: BackendSettingsDict
-    qwen: BackendSettingsDict
-    yi: BackendSettingsDict
-    zhipuai: BackendSettingsDict
-    baichuan: BackendSettingsDict
-    stepfun: BackendSettingsDict
-    xai: BackendSettingsDict
+    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]
+    gemini: NotRequired[BackendSettingsDict]
+    groq: NotRequired[BackendSettingsDict]
+    local: NotRequired[BackendSettingsDict]
+    minimax: NotRequired[BackendSettingsDict]
+    mistral: NotRequired[BackendSettingsDict]
+    moonshot: NotRequired[BackendSettingsDict]
+    openai: NotRequired[BackendSettingsDict]
+    qwen: NotRequired[BackendSettingsDict]
+    yi: NotRequired[BackendSettingsDict]
+    zhipuai: NotRequired[BackendSettingsDict]
+    baichuan: NotRequired[BackendSettingsDict]
+    stepfun: NotRequired[BackendSettingsDict]
+    xai: NotRequired[BackendSettingsDict]

{vectorvein-0.2.11.dist-info → vectorvein-0.2.13.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: vectorvein
-Version: 0.2.11
+Version: 0.2.13
 Summary: VectorVein python SDK
 Author-Email: Anderson <andersonby@163.com>
 License: MIT

{vectorvein-0.2.11.dist-info → vectorvein-0.2.13.dist-info}/RECORD

@@ -1,6 +1,6 @@
-vectorvein-0.2.11.dist-info/METADATA,sha256=gSQgVDuAhXDNgYkTNflojyYrSZudUpV76aDOSrNsb_g,4414
-vectorvein-0.2.11.dist-info/WHEEL,sha256=thaaA2w1JzcGC48WYufAs8nrYZjJm8LqNfnXFOFyCC4,90
-vectorvein-0.2.11.dist-info/entry_points.txt,sha256=6OYgBcLyFCUgeqLgnvMyOJxPCWzgy7se4rLPKtNonMs,34
+vectorvein-0.2.13.dist-info/METADATA,sha256=sEZT2h2XwDUIkIvdKzXomeSuniLUSr-HWlj88MmHRew,4414
+vectorvein-0.2.13.dist-info/WHEEL,sha256=thaaA2w1JzcGC48WYufAs8nrYZjJm8LqNfnXFOFyCC4,90
+vectorvein-0.2.13.dist-info/entry_points.txt,sha256=6OYgBcLyFCUgeqLgnvMyOJxPCWzgy7se4rLPKtNonMs,34
 vectorvein/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 vectorvein/api/__init__.py,sha256=lfY-XA46fgD2iIZTU0VYP8i07AwA03Egj4Qua0vUKrQ,738
 vectorvein/api/client.py,sha256=xF-leKDQzVyyy9FnIRaz0k4eElYW1XbbzeRLcpnyk90,33047
@@ -18,7 +18,7 @@ vectorvein/chat_clients/minimax_client.py,sha256=YOILWcsHsN5tihLTMbKJIyJr9TJREMI
 vectorvein/chat_clients/mistral_client.py,sha256=1aKSylzBDaLYcFnaBIL4-sXSzWmXfBeON9Q0rq-ziWw,534
 vectorvein/chat_clients/moonshot_client.py,sha256=gbu-6nGxx8uM_U2WlI4Wus881rFRotzHtMSoYOcruGU,526
 vectorvein/chat_clients/openai_client.py,sha256=Nz6tV45pWcsOupxjnsRsGTicbQNJWIZyxuJoJ5DGMpg,527
-vectorvein/chat_clients/openai_compatible_client.py,sha256=9ldl6TPinXLLH1kPtP-pVYO2D-VBy_s1Gf-XNNdd3CY,48498
+vectorvein/chat_clients/openai_compatible_client.py,sha256=L8SXCRA7OO_eXh6b-oya8k6XZZJ-j8SrJSvy8BFhCgs,48498
 vectorvein/chat_clients/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 vectorvein/chat_clients/qwen_client.py,sha256=-ryh-m9PgsO0fc4ulcCmPTy1155J8YUy15uPoJQOHA0,513
 vectorvein/chat_clients/stepfun_client.py,sha256=zsD2W5ahmR4DD9cqQTXmJr3txrGuvxbRWhFlRdwNijI,519
@@ -28,15 +28,15 @@ vectorvein/chat_clients/yi_client.py,sha256=RNf4CRuPJfixrwLZ3-DEc3t25QDe1mvZeb9s
 vectorvein/chat_clients/zhipuai_client.py,sha256=Ys5DSeLCuedaDXr3PfG1EW2zKXopt-awO2IylWSwY0s,519
 vectorvein/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 vectorvein/server/token_server.py,sha256=36F9PKSNOX8ZtYBXY_l-76GQTpUSmQ2Y8EMy1H7wtdQ,1353
-vectorvein/settings/__init__.py,sha256=ecGyrE_6YfX9z6Igb1rDCu1Q-qMTcVozWF3WEl_hiKA,4871
+vectorvein/settings/__init__.py,sha256=oBOLwG61RN9dq012Bgq3OfZyxyrJ5j0BYNmHbThhpjo,8097
 vectorvein/settings/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 vectorvein/types/__init__.py,sha256=ie7H3rTMq_Fg836vOmy96m3wzjDkqfekQecPXXEDbcM,3005
 vectorvein/types/defaults.py,sha256=VrkQoyHqC_eK3g1b6egpPYLLo0ltwMHqxDscCX4y-N0,27417
 vectorvein/types/enums.py,sha256=7KTJSVtQueImmbr1fSwv3rQVtc0RyMWXJmoE2tDOaso,1667
 vectorvein/types/exception.py,sha256=gnW4GnJ76jND6UGnodk9xmqkcbeS7Cz2rvncA2HpD5E,69
-vectorvein/types/llm_parameters.py,sha256=8ot02N3PS794bJTc0jlPk_XoXDGWQDTxyyF-1NmqVn0,6103
+vectorvein/types/llm_parameters.py,sha256=4SxDbJKVb9oGYymyxQtNZ66YZmUQd9_CpYYg81_Inkk,7650
 vectorvein/types/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-vectorvein/types/settings.py,sha256=hzLtE7ryyIIGxcExAsqh9sMNKP8-r4WczhYWIBNm-uM,2880
+vectorvein/types/settings.py,sha256=096mcaqtuZYCeJQYzF4ILqrLp_z5-upJAxwLrRZ5WQk,4165
 vectorvein/utilities/media_processing.py,sha256=7KtbLFzOYIn1e9QTN9G6C76NH8CBlV9kfAgiRKEIeXY,6263
 vectorvein/utilities/rate_limiter.py,sha256=dwolIUVw2wP83Odqpx0AAaE77de1GzxkYDGH4tM_u_4,10300
 vectorvein/utilities/retry.py,sha256=6KFS9R2HdhqM3_9jkjD4F36ZSpEx2YNFGOVlpOsUetM,2208
@@ -61,4 +61,4 @@ vectorvein/workflow/nodes/vector_db.py,sha256=t6I17q6iR3yQreiDHpRrksMdWDPIvgqJs0
 vectorvein/workflow/nodes/video_generation.py,sha256=qmdg-t_idpxq1veukd-jv_ChICMOoInKxprV9Z4Vi2w,4118
 vectorvein/workflow/nodes/web_crawlers.py,sha256=LsqomfXfqrXfHJDO1cl0Ox48f4St7X_SL12DSbAMSOw,5415
 vectorvein/workflow/utils/json_to_code.py,sha256=F7dhDy8kGc8ndOeihGLRLGFGlquoxVlb02ENtxnQ0C8,5914
-vectorvein-0.2.11.dist-info/RECORD,,
+vectorvein-0.2.13.dist-info/RECORD,,