vectorvein 0.1.27__tar.gz → 0.1.29__tar.gz

{vectorvein-0.1.27 → vectorvein-0.1.29}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: vectorvein
-Version: 0.1.27
+Version: 0.1.29
 Summary: Default template for PDM package
 Author-Email: Anderson <andersonby@163.com>
 License: MIT

{vectorvein-0.1.27 → vectorvein-0.1.29}/pyproject.toml

@@ -16,7 +16,7 @@ description = "Default template for PDM package"
 name = "vectorvein"
 readme = "README.md"
 requires-python = ">=3.10"
-version = "0.1.27"
+version = "0.1.29"
 
 [project.license]
 text = "MIT"

vectorvein-0.1.29/src/vectorvein/chat_clients/__init__.py

@@ -0,0 +1,522 @@
+# @Author: Bi Ying
+# @Date:   2024-07-26 14:48:55
+import httpx
+from typing import overload, Literal
+
+from .base_client import BaseChatClient, BaseAsyncChatClient
+
+from .yi_client import YiChatClient, AsyncYiChatClient
+from .groq_client import GroqChatClient, AsyncGroqChatClient
+from .qwen_client import QwenChatClient, AsyncQwenChatClient
+from .local_client import LocalChatClient, AsyncLocalChatClient
+from .gemini_client import GeminiChatClient, AsyncGeminiChatClient
+from .openai_client import OpenAIChatClient, AsyncOpenAIChatClient
+from .zhipuai_client import ZhiPuAIChatClient, AsyncZhiPuAIChatClient
+from .minimax_client import MiniMaxChatClient, AsyncMiniMaxChatClient
+from .mistral_client import MistralChatClient, AsyncMistralChatClient
+from .stepfun_client import StepFunChatClient, AsyncStepFunChatClient
+from .baichuan_client import BaichuanChatClient, AsyncBaichuanChatClient
+from .moonshot_client import MoonshotChatClient, AsyncMoonshotChatClient
+from .deepseek_client import DeepSeekChatClient, AsyncDeepSeekChatClient
+
+from ..types import defaults as defs
+from ..types.enums import BackendType, ContextLengthControlType
+from .anthropic_client import AnthropicChatClient, AsyncAnthropicChatClient
+from .utils import format_messages, get_token_counts, get_message_token_counts, ToolCallContentProcessor
+
+# 后端映射
+BackendMap = {
+    "sync": {
+        BackendType.Anthropic: AnthropicChatClient,
+        BackendType.DeepSeek: DeepSeekChatClient,
+        BackendType.Gemini: GeminiChatClient,
+        BackendType.Groq: GroqChatClient,
+        BackendType.Local: LocalChatClient,
+        BackendType.MiniMax: MiniMaxChatClient,
+        BackendType.Mistral: MistralChatClient,
+        BackendType.Moonshot: MoonshotChatClient,
+        BackendType.OpenAI: OpenAIChatClient,
+        BackendType.Qwen: QwenChatClient,
+        BackendType.Yi: YiChatClient,
+        BackendType.ZhiPuAI: ZhiPuAIChatClient,
+        BackendType.Baichuan: BaichuanChatClient,
+        BackendType.StepFun: StepFunChatClient,
+    },
+    "async": {
+        BackendType.Anthropic: AsyncAnthropicChatClient,
+        BackendType.DeepSeek: AsyncDeepSeekChatClient,
+        BackendType.Gemini: AsyncGeminiChatClient,
+        BackendType.Groq: AsyncGroqChatClient,
+        BackendType.Local: AsyncLocalChatClient,
+        BackendType.MiniMax: AsyncMiniMaxChatClient,
+        BackendType.Mistral: AsyncMistralChatClient,
+        BackendType.Moonshot: AsyncMoonshotChatClient,
+        BackendType.OpenAI: AsyncOpenAIChatClient,
+        BackendType.Qwen: AsyncQwenChatClient,
+        BackendType.Yi: AsyncYiChatClient,
+        BackendType.ZhiPuAI: AsyncZhiPuAIChatClient,
+        BackendType.Baichuan: AsyncBaichuanChatClient,
+        BackendType.StepFun: AsyncStepFunChatClient,
+    },
+}
+
+
+@overload
+def create_chat_client(
+    backend: Literal[BackendType.Anthropic],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.Client | None = None,
+    **kwargs,
+) -> AnthropicChatClient: ...
+
+
+@overload
+def create_chat_client(
+    backend: Literal[BackendType.DeepSeek],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.Client | None = None,
+    **kwargs,
+) -> DeepSeekChatClient: ...
+
+
+@overload
+def create_chat_client(
+    backend: Literal[BackendType.Gemini],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.Client | None = None,
+    **kwargs,
+) -> GeminiChatClient: ...
+
+
+@overload
+def create_chat_client(
+    backend: Literal[BackendType.Groq],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.Client | None = None,
+    **kwargs,
+) -> GroqChatClient: ...
+
+
+@overload
+def create_chat_client(
+    backend: Literal[BackendType.Local],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.Client | None = None,
+    **kwargs,
+) -> LocalChatClient: ...
+
+
+@overload
+def create_chat_client(
+    backend: Literal[BackendType.MiniMax],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.Client | None = None,
+    **kwargs,
+) -> MiniMaxChatClient: ...
+
+
+@overload
+def create_chat_client(
+    backend: Literal[BackendType.Mistral],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.Client | None = None,
+    **kwargs,
+) -> MistralChatClient: ...
+
+
+@overload
+def create_chat_client(
+    backend: Literal[BackendType.Moonshot],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.Client | None = None,
+    **kwargs,
+) -> MoonshotChatClient: ...
+
+
+@overload
+def create_chat_client(
+    backend: Literal[BackendType.OpenAI],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.Client | None = None,
+    **kwargs,
+) -> OpenAIChatClient: ...
+
+
+@overload
+def create_chat_client(
+    backend: Literal[BackendType.Qwen],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.Client | None = None,
+    **kwargs,
+) -> QwenChatClient: ...
+
+
+@overload
+def create_chat_client(
+    backend: Literal[BackendType.Yi],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.Client | None = None,
+    **kwargs,
+) -> YiChatClient: ...
+
+
+@overload
+def create_chat_client(
+    backend: Literal[BackendType.ZhiPuAI],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.Client | None = None,
+    **kwargs,
+) -> ZhiPuAIChatClient: ...
+
+
+@overload
+def create_chat_client(
+    backend: Literal[BackendType.Baichuan],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.Client | None = None,
+    **kwargs,
+) -> BaichuanChatClient: ...
+
+
+@overload
+def create_chat_client(
+    backend: Literal[BackendType.StepFun],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.Client | None = None,
+    **kwargs,
+) -> StepFunChatClient: ...
+
+
+def create_chat_client(
+    backend: BackendType,
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.Client | None = None,
+    **kwargs,
+) -> BaseChatClient:
+    if backend not in BackendMap["sync"]:
+        raise ValueError(f"Unsupported backend: {backend}")
+
+    ClientClass = BackendMap["sync"][backend]
+    if model is None:
+        model = ClientClass.DEFAULT_MODEL
+    return ClientClass(
+        model=model,
+        stream=stream,
+        temperature=temperature,
+        context_length_control=context_length_control,
+        random_endpoint=random_endpoint,
+        endpoint_id=endpoint_id,
+        http_client=http_client,
+        **kwargs,
+    )
+
+
+@overload
+def create_async_chat_client(
+    backend: Literal[BackendType.Anthropic],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.AsyncClient | None = None,
+    **kwargs,
+) -> AsyncAnthropicChatClient: ...
+
+
+@overload
+def create_async_chat_client(
+    backend: Literal[BackendType.DeepSeek],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.AsyncClient | None = None,
+    **kwargs,
+) -> AsyncDeepSeekChatClient: ...
+
+
+@overload
+def create_async_chat_client(
+    backend: Literal[BackendType.Gemini],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.AsyncClient | None = None,
+    **kwargs,
+) -> AsyncGeminiChatClient: ...
+
+
+@overload
+def create_async_chat_client(
+    backend: Literal[BackendType.Groq],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.AsyncClient | None = None,
+    **kwargs,
+) -> AsyncGroqChatClient: ...
+
+
+@overload
+def create_async_chat_client(
+    backend: Literal[BackendType.Local],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.AsyncClient | None = None,
+    **kwargs,
+) -> AsyncLocalChatClient: ...
+
+
+@overload
+def create_async_chat_client(
+    backend: Literal[BackendType.MiniMax],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.AsyncClient | None = None,
+    **kwargs,
+) -> AsyncMiniMaxChatClient: ...
+
+
+@overload
+def create_async_chat_client(
+    backend: Literal[BackendType.Mistral],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.AsyncClient | None = None,
+    **kwargs,
+) -> AsyncMistralChatClient: ...
+
+
+@overload
+def create_async_chat_client(
+    backend: Literal[BackendType.Moonshot],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.AsyncClient | None = None,
+    **kwargs,
+) -> AsyncMoonshotChatClient: ...
+
+
+@overload
+def create_async_chat_client(
+    backend: Literal[BackendType.OpenAI],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.AsyncClient | None = None,
+    **kwargs,
+) -> AsyncOpenAIChatClient: ...
+
+
+@overload
+def create_async_chat_client(
+    backend: Literal[BackendType.Qwen],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.AsyncClient | None = None,
+    **kwargs,
+) -> AsyncQwenChatClient: ...
+
+
+@overload
+def create_async_chat_client(
+    backend: Literal[BackendType.Yi],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.AsyncClient | None = None,
+    **kwargs,
+) -> AsyncYiChatClient: ...
+
+
+@overload
+def create_async_chat_client(
+    backend: Literal[BackendType.ZhiPuAI],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.AsyncClient | None = None,
+    **kwargs,
+) -> AsyncZhiPuAIChatClient: ...
+
+
+@overload
+def create_async_chat_client(
+    backend: Literal[BackendType.Baichuan],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.AsyncClient | None = None,
+    **kwargs,
+) -> AsyncBaichuanChatClient: ...
+
+
+@overload
+def create_async_chat_client(
+    backend: Literal[BackendType.StepFun],
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.AsyncClient | None = None,
+    **kwargs,
+) -> AsyncStepFunChatClient: ...
+
+
+def create_async_chat_client(
+    backend: BackendType,
+    model: str | None = None,
+    stream: bool = False,
+    temperature: float = 0.7,
+    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
+    random_endpoint: bool = True,
+    endpoint_id: str = "",
+    http_client: httpx.AsyncClient | None = None,
+    **kwargs,
+) -> BaseAsyncChatClient:
+    if backend not in BackendMap["async"]:
+        raise ValueError(f"Unsupported backend: {backend}")
+
+    ClientClass = BackendMap["async"][backend]
+    if model is None:
+        model = ClientClass.DEFAULT_MODEL
+    return ClientClass(
+        model=model,
+        stream=stream,
+        temperature=temperature,
+        context_length_control=context_length_control,
+        random_endpoint=random_endpoint,
+        endpoint_id=endpoint_id,
+        http_client=http_client,
+        **kwargs,
+    )
+
+
+__all__ = [
+    "BackendType",
+    "format_messages",
+    "get_token_counts",
+    "create_chat_client",
+    "create_async_chat_client",
+    "get_message_token_counts",
+    "ToolCallContentProcessor",
+]

{vectorvein-0.1.27 → vectorvein-0.1.29}/src/vectorvein/chat_clients/openai_compatible_client.py

@@ -169,10 +169,10 @@ class OpenAICompatibleChatClient(BaseChatClient):
             max_output_tokens = self.model_setting.max_output_tokens
             token_counts = get_message_token_counts(messages=messages, tools=tools, model=self.model_setting.id)
             if max_output_tokens is not None:
-                max_tokens = self.model_setting.context_length - token_counts
+                max_tokens = self.model_setting.context_length - token_counts - 64
                 max_tokens = min(max(max_tokens, 1), max_output_tokens)
             else:
-                max_tokens = self.model_setting.context_length - token_counts
+                max_tokens = self.model_setting.context_length - token_counts - 64
 
         if response_format and self.model_setting.response_format_available:
             self.response_format = {"response_format": response_format}
@@ -396,10 +396,10 @@ class AsyncOpenAICompatibleChatClient(BaseAsyncChatClient):
             max_output_tokens = self.model_setting.max_output_tokens
             token_counts = get_message_token_counts(messages=messages, tools=tools, model=self.model_setting.id)
             if max_output_tokens is not None:
-                max_tokens = self.model_setting.context_length - token_counts
+                max_tokens = self.model_setting.context_length - token_counts - 64
                 max_tokens = min(max(max_tokens, 1), max_output_tokens)
             else:
-                max_tokens = self.model_setting.context_length - token_counts
+                max_tokens = self.model_setting.context_length - token_counts - 64
 
         if self.stream:
             stream_response: AsyncStream[ChatCompletionChunk] = await self.raw_client.chat.completions.create(

vectorvein-0.1.29/src/vectorvein/chat_clients/stepfun_client.py

@@ -0,0 +1,15 @@
+# @Author: Bi Ying
+# @Date:   2024-07-26 14:48:55
+from ..types.enums import BackendType
+from ..types.defaults import STEPFUN_DEFAULT_MODEL
+from .openai_compatible_client import OpenAICompatibleChatClient, AsyncOpenAICompatibleChatClient
+
+
+class StepFunChatClient(OpenAICompatibleChatClient):
+    DEFAULT_MODEL = STEPFUN_DEFAULT_MODEL
+    BACKEND_NAME = BackendType.StepFun
+
+
+class AsyncStepFunChatClient(AsyncOpenAICompatibleChatClient):
+    DEFAULT_MODEL = STEPFUN_DEFAULT_MODEL
+    BACKEND_NAME = BackendType.StepFun

{vectorvein-0.1.27 → vectorvein-0.1.29}/src/vectorvein/chat_clients/utils.py

@@ -195,6 +195,31 @@ def get_token_counts(text: str | dict, model: str = "") -> int:
         return len(deepseek_tokenizer.encode(text))
     elif model.startswith("qwen"):
         return len(qwen_tokenizer.encode(text))
+    elif model.startswith("stepfun"):
+        model_setting = settings.moonshot.models[model]
+        if len(model_setting.endpoints) == 0:
+            return len(chatgpt_encoding.encode(text))
+        endpoint_id = model_setting.endpoints[0]
+        endpoint = settings.get_endpoint(endpoint_id)
+        tokenize_url = "https://api.stepfun.com/v1/token/count"
+        headers = {"Content-Type": "application/json", "Authorization": f"Bearer {endpoint.api_key}"}
+        request_body = {
+            "model": model,
+            "messages": [
+                {"role": "user", "content": text},
+            ],
+        }
+        _, response = (
+            Retry(httpx.post)
+            .args(url=tokenize_url, headers=headers, json=request_body, timeout=None)
+            .retry_times(5)
+            .sleep_time(10)
+            .run()
+        )
+        if response is None:
+            return 1000
+        result = response.json()
+        return result["data"]["total_tokens"]
     else:
         return len(chatgpt_encoding.encode(text))
 

{vectorvein-0.1.27 → vectorvein-0.1.29}/src/vectorvein/settings/__init__.py

@@ -27,6 +27,7 @@ class Settings(BaseModel):
     yi: BackendSettings = Field(default_factory=BackendSettings, description="Yi models settings.")
     zhipuai: BackendSettings = Field(default_factory=BackendSettings, description="Zhipuai models settings.")
     baichuan: BackendSettings = Field(default_factory=BackendSettings, description="Baichuan models settings.")
+    stepfun: BackendSettings = Field(default_factory=BackendSettings, description="StepFun models settings.")
 
     def __init__(self, **data):
         model_types = {
@@ -43,6 +44,7 @@ class Settings(BaseModel):
             "yi": defs.YI_MODELS,
             "zhipuai": defs.ZHIPUAI_MODELS,
             "baichuan": defs.BAICHUAN_MODELS,
+            "stepfun": defs.STEPFUN_MODELS,
         }
 
         for model_type, default_models in model_types.items():
@@ -62,7 +64,7 @@ class Settings(BaseModel):
         for endpoint in self.endpoints:
             if endpoint.id == endpoint_id:
                 return endpoint
-        return EndpointSetting()
+        raise ValueError(f"Endpoint {endpoint_id} not found.")
 
     def get_backend(self, backend: BackendType) -> BackendSettings:
         return getattr(self, backend.value.lower())

{vectorvein-0.1.27 → vectorvein-0.1.29}/src/vectorvein/types/defaults.py

@@ -588,3 +588,98 @@ GEMINI_MODELS = {
         "native_multimodal": True,
     },
 }
+
+# 百度文心一言 ERNIE 模型
+ERNIE_DEFAULT_MODEL = "ernie-lite"
+ERNIE_MODELS = {
+    "ernie-lite": {
+        "id": "ernie-lite",
+        "context_length": 6144,
+        "max_output_tokens": 2048,
+        "function_call_available": False,
+        "response_format_available": False,
+    },
+    "ernie-speed": {
+        "id": "ernie-speed",
+        "context_length": 126976,
+        "max_output_tokens": 4096,
+        "function_call_available": False,
+        "response_format_available": False,
+    },
+    "ernie-speed-pro-128k": {
+        "id": "ernie-speed-pro-128k",
+        "context_length": 126976,
+        "max_output_tokens": 4096,
+        "function_call_available": False,
+        "response_format_available": False,
+    },
+    "ernie-4.0-8k-latest": {
+        "id": "ernie-4.0-8k-latest",
+        "context_length": 5120,
+        "max_output_tokens": 2048,
+        "function_call_available": False,
+        "response_format_available": True,
+    },
+    "ernie-4.0-turbo-8k": {
+        "id": "ernie-4.0-turbo-8k",
+        "context_length": 5120,
+        "max_output_tokens": 2048,
+        "function_call_available": False,
+        "response_format_available": True,
+    },
+}
+
+
+STEPFUN_DEFAULT_MODEL = "step-1-8k"
+STEPFUN_MODELS = {
+    "step-1-8k": {
+        "id": "step-1-8k",
+        "context_length": 8192,
+        "function_call_available": True,
+        "response_format_available": True,
+    },
+    "step-1-32k": {
+        "id": "step-1-32k",
+        "context_length": 32000,
+        "function_call_available": True,
+        "response_format_available": True,
+    },
+    "step-1-128k": {
+        "id": "step-1-128k",
+        "context_length": 128000,
+        "function_call_available": True,
+        "response_format_available": True,
+    },
+    "step-1-256k": {
+        "id": "step-1-256k",
+        "context_length": 256000,
+        "function_call_available": True,
+        "response_format_available": True,
+    },
+    "step-2-16k": {
+        "id": "step-2-16k",
+        "context_length": 16384,
+        "function_call_available": True,
+        "response_format_available": True,
+    },
+    "step-1-flash": {
+        "id": "step-1-flash",
+        "context_length": 8192,
+        "function_call_available": True,
+        "response_format_available": True,
+    },
+    "step-1v-8k": {
+        "id": "step-1v-8k",
+        "context_length": 8192,
+        "function_call_available": False,
+        "response_format_available": False,
+        "native_multimodal": True,
+    },
+    "step-1v-32k": {
+        "id": "step-1v-32k",
+        "context_length": 32768,
+        "function_call_available": False,
+        "response_format_available": False,
+        "native_multimodal": True,
+    },
+}

{vectorvein-0.1.27 → vectorvein-0.1.29}/src/vectorvein/types/enums.py

@@ -47,6 +47,9 @@ class BackendType(str, Enum):
     # Baichuan
     Baichuan = "baichuan"
 
+    # StepFun
+    StepFun = "stepfun"
+
     def __repr__(self):
         """Get a string representation."""
         return f'"{self.value}"'

vectorvein-0.1.27/src/vectorvein/chat_clients/__init__.py

@@ -1,130 +0,0 @@
-# @Author: Bi Ying
-# @Date:   2024-07-26 14:48:55
-import httpx
-
-from .base_client import BaseChatClient, BaseAsyncChatClient
-
-from .yi_client import YiChatClient, AsyncYiChatClient
-from .groq_client import GroqChatClient, AsyncGroqChatClient
-from .qwen_client import QwenChatClient, AsyncQwenChatClient
-from .local_client import LocalChatClient, AsyncLocalChatClient
-from .gemini_client import GeminiChatClient, AsyncGeminiChatClient
-from .openai_client import OpenAIChatClient, AsyncOpenAIChatClient
-from .zhipuai_client import ZhiPuAIChatClient, AsyncZhiPuAIChatClient
-from .minimax_client import MiniMaxChatClient, AsyncMiniMaxChatClient
-from .mistral_client import MistralChatClient, AsyncMistralChatClient
-from .baichuan_client import BaichuanChatClient, AsyncBaichuanChatClient
-from .moonshot_client import MoonshotChatClient, AsyncMoonshotChatClient
-from .deepseek_client import DeepSeekChatClient, AsyncDeepSeekChatClient
-
-from ..types import defaults as defs
-from ..types.enums import BackendType, ContextLengthControlType
-from .anthropic_client import AnthropicChatClient, AsyncAnthropicChatClient
-from .utils import format_messages, get_token_counts, get_message_token_counts, ToolCallContentProcessor
-
-
-BackendMap = {
-    "sync": {
-        BackendType.Anthropic: AnthropicChatClient,
-        BackendType.DeepSeek: DeepSeekChatClient,
-        BackendType.Gemini: GeminiChatClient,
-        BackendType.Groq: GroqChatClient,
-        BackendType.Local: LocalChatClient,
-        BackendType.MiniMax: MiniMaxChatClient,
-        BackendType.Mistral: MistralChatClient,
-        BackendType.Moonshot: MoonshotChatClient,
-        BackendType.OpenAI: OpenAIChatClient,
-        BackendType.Qwen: QwenChatClient,
-        BackendType.Yi: YiChatClient,
-        BackendType.ZhiPuAI: ZhiPuAIChatClient,
-        BackendType.Baichuan: BaichuanChatClient,
-    },
-    "async": {
-        BackendType.Anthropic: AsyncAnthropicChatClient,
-        BackendType.DeepSeek: AsyncDeepSeekChatClient,
-        BackendType.Gemini: AsyncGeminiChatClient,
-        BackendType.Groq: AsyncGroqChatClient,
-        BackendType.Local: AsyncLocalChatClient,
-        BackendType.MiniMax: AsyncMiniMaxChatClient,
-        BackendType.Mistral: AsyncMistralChatClient,
-        BackendType.Moonshot: AsyncMoonshotChatClient,
-        BackendType.OpenAI: AsyncOpenAIChatClient,
-        BackendType.Qwen: AsyncQwenChatClient,
-        BackendType.Yi: AsyncYiChatClient,
-        BackendType.ZhiPuAI: AsyncZhiPuAIChatClient,
-        BackendType.Baichuan: AsyncBaichuanChatClient,
-    },
-}
-
-
-def create_chat_client(
-    backend: BackendType,
-    model: str | None = None,
-    stream: bool = False,
-    temperature: float = 0.7,
-    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
-    random_endpoint: bool = True,
-    endpoint_id: str = "",
-    http_client: httpx.Client | None = None,
-    **kwargs,
-) -> BaseChatClient:
-    if backend.lower() not in BackendMap["sync"]:
-        raise ValueError(f"Unsupported backend: {backend}")
-    else:
-        backend_key = backend.lower()
-
-    ClientClass = BackendMap["sync"][backend_key]
-    if model is None:
-        model = ClientClass.DEFAULT_MODEL
-    return BackendMap["sync"][backend_key](
-        model=model,
-        stream=stream,
-        temperature=temperature,
-        context_length_control=context_length_control,
-        random_endpoint=random_endpoint,
-        endpoint_id=endpoint_id,
-        http_client=http_client,
-        **kwargs,
-    )
-
-
-def create_async_chat_client(
-    backend: BackendType,
-    model: str | None = None,
-    stream: bool = False,
-    temperature: float = 0.7,
-    context_length_control: ContextLengthControlType = defs.CONTEXT_LENGTH_CONTROL,
-    random_endpoint: bool = True,
-    endpoint_id: str = "",
-    http_client: httpx.AsyncClient | None = None,
-    **kwargs,
-) -> BaseAsyncChatClient:
-    if backend.lower() not in BackendMap["async"]:
-        raise ValueError(f"Unsupported backend: {backend}")
-    else:
-        backend_key = backend.lower()
-
-    ClientClass = BackendMap["async"][backend_key]
-    if model is None:
-        model = ClientClass.DEFAULT_MODEL
-    return BackendMap["async"][backend_key](
-        model=model,
-        stream=stream,
-        temperature=temperature,
-        context_length_control=context_length_control,
-        random_endpoint=random_endpoint,
-        endpoint_id=endpoint_id,
-        http_client=http_client,
-        **kwargs,
-    )
-
-
-__all__ = [
-    "BackendType",
-    "format_messages",
-    "get_token_counts",
-    "create_chat_client",
-    "create_async_chat_client",
-    "get_message_token_counts",
-    "ToolCallContentProcessor",
-]