vectorvein 0.2.11__tar.gz → 0.2.13__tar.gz

{vectorvein-0.2.11 → vectorvein-0.2.13}/PKG-INFO

@@ -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 → vectorvein-0.2.13}/pyproject.toml

@@ -18,7 +18,7 @@ description = "VectorVein python SDK"
 name = "vectorvein"
 readme = "README.md"
 requires-python = ">=3.10"
-version = "0.2.11"
+version = "0.2.13"
 
 [project.license]
 text = "MIT"

{vectorvein-0.2.11 → vectorvein-0.2.13}/src/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-0.2.11 → vectorvein-0.2.13}/src/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-0.2.11 → vectorvein-0.2.13}/src/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-0.2.13/src/vectorvein/types/settings.py

@@ -0,0 +1,130 @@
+from typing import Dict, List, Optional, Union, Literal
+from typing_extensions import TypedDict, NotRequired  # Required by pydantic under Python < 3.12
+
+
+class RedisConfigDict(TypedDict):
+    """TypedDict representing the RedisConfig structure."""
+
+    host: str
+    port: int
+    db: int
+
+
+class DiskCacheConfigDict(TypedDict):
+    """TypedDict representing the DiskCacheConfig structure."""
+
+    cache_dir: str
+
+
+class RateLimitConfigDict(TypedDict):
+    """TypedDict representing the RateLimitConfig structure."""
+
+    enabled: bool
+    backend: Literal["memory", "redis", "diskcache"]
+    redis: NotRequired[RedisConfigDict]
+    diskcache: NotRequired[DiskCacheConfigDict]
+    default_rpm: int
+    default_tpm: int
+
+
+class ServerDict(TypedDict):
+    """TypedDict representing the Server structure."""
+
+    host: str
+    port: int
+    url: NotRequired[str]
+
+
+class EndpointOptionDict(TypedDict):
+    """TypedDict representing the model endpoint option structure."""
+
+    endpoint_id: str
+    model_id: str
+    rpm: NotRequired[int]
+    tpm: NotRequired[int]
+    concurrent_requests: NotRequired[int]
+
+
+class ModelConfigDict(TypedDict):
+    """TypedDict representing the model configuration structure."""
+
+    id: str
+    endpoints: List[Union[str, EndpointOptionDict]]
+    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):
+    """TypedDict representing the BackendSettings structure."""
+
+    models: Dict[str, ModelConfigDict]
+
+
+class EndpointSettingDict(TypedDict):
+    """TypedDict representing the EndpointSetting structure."""
+
+    id: 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: 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/src/vectorvein/types/settings.py

@@ -1,105 +0,0 @@
-from typing import Dict, List, Optional, Union, Literal
-from typing_extensions import TypedDict, NotRequired  # Required by pydantic under Python < 3.12
-
-
-class RedisConfigDict(TypedDict):
-    """TypedDict representing the RedisConfig structure."""
-
-    host: str
-    port: int
-    db: int
-
-
-class DiskCacheConfigDict(TypedDict):
-    """TypedDict representing the DiskCacheConfig structure."""
-
-    cache_dir: str
-
-
-class RateLimitConfigDict(TypedDict):
-    """TypedDict representing the RateLimitConfig structure."""
-
-    enabled: bool
-    backend: Literal["memory", "redis", "diskcache"]
-    redis: Optional[RedisConfigDict]
-    diskcache: Optional[DiskCacheConfigDict]
-    default_rpm: int
-    default_tpm: int
-
-
-class ServerDict(TypedDict):
-    """TypedDict representing the Server structure."""
-
-    host: str
-    port: int
-    url: Optional[str]
-
-
-class EndpointOptionDict(TypedDict):
-    """TypedDict representing the model endpoint option structure."""
-
-    endpoint_id: str
-    model_id: str
-    rpm: NotRequired[int]
-    tpm: NotRequired[int]
-    concurrent_requests: NotRequired[int]
-
-
-class ModelConfigDict(TypedDict):
-    """TypedDict representing the model configuration structure."""
-
-    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]
-
-
-class BackendSettingsDict(TypedDict):
-    """TypedDict representing the BackendSettings structure."""
-
-    models: Dict[str, ModelConfigDict]
-
-
-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]
-
-
-class SettingsDict(TypedDict):
-    """TypedDict representing the expected structure of the settings dictionary."""
-
-    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