mirror of
https://github.com/zhenxun-org/zhenxun_bot.git
synced 2026-10-18 10:40:03 +08:00
✨ feat(llm): 实现LLM服务模块，支持多提供商统一接口和高级功能 (#1923)
* ✨ feat(llm): 实现LLM服务模块，支持多提供商统一接口和高级功能 * 🎨 Ruff * ✨ Config配置类支持BaseModel存储 * 🎨 代码格式化 * 🎨 代码格式化 * 🎨 格式化代码 * ✨ feat(llm): 添加 AI 对话历史管理 * ✨ feat(llmConfig): 引入 LLM 配置模型及管理功能 * 🎨 Ruff --------- Co-authored-by: fccckaug <xxxmio123123@gmail.com> Co-authored-by: HibiKier <45528451+HibiKier@users.noreply.github.com> Co-authored-by: HibiKier <775757368@qq.com> Co-authored-by: fccckaug <xxxmcsmiomio3@gmail.com> Co-authored-by: webjoin111 <455457521@qq.com>
This commit is contained in:
Rumio
co-authored by
fccckaug
HibiKier
HibiKier
fccckaug
webjoin111
fccckaug
HibiKier
HibiKier
fccckaug
webjoin111
parent
14f5842f10
commit
a020ea5c87
@@ -0,0 +1,35 @@
|
||||
"""
|
||||
LLM 配置模块
|
||||
|
||||
提供生成配置、预设配置和配置验证功能。
|
||||
"""
|
||||
|
||||
from .generation import (
|
||||
LLMGenerationConfig,
|
||||
ModelConfigOverride,
|
||||
apply_api_specific_mappings,
|
||||
create_generation_config_from_kwargs,
|
||||
validate_override_params,
|
||||
)
|
||||
from .presets import CommonOverrides
|
||||
from .providers import (
|
||||
LLMConfig,
|
||||
get_llm_config,
|
||||
register_llm_configs,
|
||||
set_default_model,
|
||||
validate_llm_config,
|
||||
)
|
||||
|
||||
__all__ = [
|
||||
"CommonOverrides",
|
||||
"LLMConfig",
|
||||
"LLMGenerationConfig",
|
||||
"ModelConfigOverride",
|
||||
"apply_api_specific_mappings",
|
||||
"create_generation_config_from_kwargs",
|
||||
"get_llm_config",
|
||||
"register_llm_configs",
|
||||
"set_default_model",
|
||||
"validate_llm_config",
|
||||
"validate_override_params",
|
||||
]
|
||||
@@ -0,0 +1,260 @@
|
||||
"""
|
||||
LLM 生成配置相关类和函数
|
||||
"""
|
||||
|
||||
from typing import Any
|
||||
|
||||
from pydantic import BaseModel, Field
|
||||
|
||||
from zhenxun.services.log import logger
|
||||
|
||||
from ..types.enums import ResponseFormat
|
||||
from ..types.exceptions import LLMErrorCode, LLMException
|
||||
|
||||
|
||||
class ModelConfigOverride(BaseModel):
|
||||
"""模型配置覆盖参数"""
|
||||
|
||||
temperature: float | None = Field(
|
||||
default=None, ge=0.0, le=2.0, description="生成温度"
|
||||
)
|
||||
max_tokens: int | None = Field(default=None, gt=0, description="最大输出token数")
|
||||
top_p: float | None = Field(default=None, ge=0.0, le=1.0, description="核采样参数")
|
||||
top_k: int | None = Field(default=None, gt=0, description="Top-K采样参数")
|
||||
frequency_penalty: float | None = Field(
|
||||
default=None, ge=-2.0, le=2.0, description="频率惩罚"
|
||||
)
|
||||
presence_penalty: float | None = Field(
|
||||
default=None, ge=-2.0, le=2.0, description="存在惩罚"
|
||||
)
|
||||
repetition_penalty: float | None = Field(
|
||||
default=None, ge=0.0, le=2.0, description="重复惩罚"
|
||||
)
|
||||
|
||||
stop: list[str] | str | None = Field(default=None, description="停止序列")
|
||||
|
||||
response_format: ResponseFormat | dict[str, Any] | None = Field(
|
||||
default=None, description="期望的响应格式"
|
||||
)
|
||||
response_mime_type: str | None = Field(
|
||||
default=None, description="响应MIME类型（Gemini专用）"
|
||||
)
|
||||
response_schema: dict[str, Any] | None = Field(
|
||||
default=None, description="JSON响应模式"
|
||||
)
|
||||
thinking_budget: float | None = Field(
|
||||
default=None, ge=0.0, le=1.0, description="思考预算"
|
||||
)
|
||||
safety_settings: dict[str, str] | None = Field(default=None, description="安全设置")
|
||||
response_modalities: list[str] | None = Field(
|
||||
default=None, description="响应模态类型"
|
||||
)
|
||||
|
||||
enable_code_execution: bool | None = Field(
|
||||
default=None, description="是否启用代码执行"
|
||||
)
|
||||
enable_grounding: bool | None = Field(
|
||||
default=None, description="是否启用信息来源关联"
|
||||
)
|
||||
enable_caching: bool | None = Field(default=None, description="是否启用响应缓存")
|
||||
|
||||
custom_params: dict[str, Any] | None = Field(default=None, description="自定义参数")
|
||||
|
||||
def to_dict(self) -> dict[str, Any]:
|
||||
"""转换为字典，排除None值"""
|
||||
result = {}
|
||||
model_data = getattr(self, "model_dump", lambda: {})()
|
||||
if not model_data:
|
||||
model_data = {}
|
||||
for field_name, _ in self.__class__.__dict__.get(
|
||||
"model_fields", {}
|
||||
).items():
|
||||
value = getattr(self, field_name, None)
|
||||
if value is not None:
|
||||
model_data[field_name] = value
|
||||
for key, value in model_data.items():
|
||||
if value is not None:
|
||||
if key == "custom_params" and isinstance(value, dict):
|
||||
result.update(value)
|
||||
else:
|
||||
result[key] = value
|
||||
return result
|
||||
|
||||
def merge_with_base_config(
|
||||
self,
|
||||
base_temperature: float | None = None,
|
||||
base_max_tokens: int | None = None,
|
||||
) -> dict[str, Any]:
|
||||
"""与基础配置合并，覆盖参数优先"""
|
||||
merged = {}
|
||||
|
||||
if base_temperature is not None:
|
||||
merged["temperature"] = base_temperature
|
||||
if base_max_tokens is not None:
|
||||
merged["max_tokens"] = base_max_tokens
|
||||
|
||||
override_dict = self.to_dict()
|
||||
merged.update(override_dict)
|
||||
|
||||
return merged
|
||||
|
||||
|
||||
class LLMGenerationConfig(ModelConfigOverride):
|
||||
"""LLM 生成配置，继承模型配置覆盖参数"""
|
||||
|
||||
def to_api_params(self, api_type: str, model_name: str) -> dict[str, Any]:
|
||||
"""转换为API参数，支持不同API类型的参数名映射"""
|
||||
_ = model_name
|
||||
params = {}
|
||||
|
||||
if self.temperature is not None:
|
||||
params["temperature"] = self.temperature
|
||||
|
||||
if self.max_tokens is not None:
|
||||
if api_type in ["gemini", "gemini_native"]:
|
||||
params["maxOutputTokens"] = self.max_tokens
|
||||
else:
|
||||
params["max_tokens"] = self.max_tokens
|
||||
|
||||
if api_type in ["gemini", "gemini_native"]:
|
||||
if self.top_k is not None:
|
||||
params["topK"] = self.top_k
|
||||
if self.top_p is not None:
|
||||
params["topP"] = self.top_p
|
||||
else:
|
||||
if self.top_k is not None:
|
||||
params["top_k"] = self.top_k
|
||||
if self.top_p is not None:
|
||||
params["top_p"] = self.top_p
|
||||
|
||||
if api_type in ["openai", "deepseek", "zhipu", "general_openai_compat"]:
|
||||
if self.frequency_penalty is not None:
|
||||
params["frequency_penalty"] = self.frequency_penalty
|
||||
if self.presence_penalty is not None:
|
||||
params["presence_penalty"] = self.presence_penalty
|
||||
|
||||
if self.repetition_penalty is not None:
|
||||
if api_type == "openai":
|
||||
logger.warning("OpenAI官方API不支持repetition_penalty参数，已忽略")
|
||||
else:
|
||||
params["repetition_penalty"] = self.repetition_penalty
|
||||
|
||||
if self.response_format is not None:
|
||||
if isinstance(self.response_format, dict):
|
||||
if api_type in ["openai", "zhipu", "deepseek", "general_openai_compat"]:
|
||||
params["response_format"] = self.response_format
|
||||
logger.debug(
|
||||
f"为 {api_type} 使用自定义 response_format: "
|
||||
f"{self.response_format}"
|
||||
)
|
||||
elif self.response_format == ResponseFormat.JSON:
|
||||
if api_type in ["openai", "zhipu", "deepseek", "general_openai_compat"]:
|
||||
params["response_format"] = {"type": "json_object"}
|
||||
logger.debug(f"为 {api_type} 启用 JSON 对象输出模式")
|
||||
elif api_type in ["gemini", "gemini_native"]:
|
||||
params["responseMimeType"] = "application/json"
|
||||
if self.response_schema:
|
||||
params["responseSchema"] = self.response_schema
|
||||
logger.debug(f"为 {api_type} 启用 JSON MIME 类型输出模式")
|
||||
|
||||
if api_type in ["gemini", "gemini_native"]:
|
||||
if (
|
||||
self.response_format != ResponseFormat.JSON
|
||||
and self.response_mime_type is not None
|
||||
):
|
||||
params["responseMimeType"] = self.response_mime_type
|
||||
logger.debug(
|
||||
f"使用显式设置的 responseMimeType: {self.response_mime_type}"
|
||||
)
|
||||
|
||||
if self.response_schema is not None and "responseSchema" not in params:
|
||||
params["responseSchema"] = self.response_schema
|
||||
if self.thinking_budget is not None:
|
||||
params["thinkingBudget"] = self.thinking_budget
|
||||
if self.safety_settings is not None:
|
||||
params["safetySettings"] = self.safety_settings
|
||||
if self.response_modalities is not None:
|
||||
params["responseModalities"] = self.response_modalities
|
||||
|
||||
if self.custom_params:
|
||||
custom_mapped = apply_api_specific_mappings(self.custom_params, api_type)
|
||||
params.update(custom_mapped)
|
||||
|
||||
logger.debug(f"为{api_type}转换配置参数: {len(params)}个参数")
|
||||
return params
|
||||
|
||||
|
||||
def validate_override_params(
|
||||
override_config: dict[str, Any] | LLMGenerationConfig | None,
|
||||
) -> LLMGenerationConfig:
|
||||
"""验证和标准化覆盖参数"""
|
||||
if override_config is None:
|
||||
return LLMGenerationConfig()
|
||||
|
||||
if isinstance(override_config, dict):
|
||||
try:
|
||||
filtered_config = {
|
||||
k: v for k, v in override_config.items() if v is not None
|
||||
}
|
||||
return LLMGenerationConfig(**filtered_config)
|
||||
except Exception as e:
|
||||
logger.warning(f"覆盖配置参数验证失败: {e}")
|
||||
raise LLMException(
|
||||
f"无效的覆盖配置参数: {e}",
|
||||
code=LLMErrorCode.CONFIGURATION_ERROR,
|
||||
cause=e,
|
||||
)
|
||||
|
||||
return override_config
|
||||
|
||||
|
||||
def apply_api_specific_mappings(
|
||||
params: dict[str, Any], api_type: str
|
||||
) -> dict[str, Any]:
|
||||
"""应用API特定的参数映射"""
|
||||
mapped_params = params.copy()
|
||||
|
||||
if api_type in ["gemini", "gemini_native"]:
|
||||
if "max_tokens" in mapped_params:
|
||||
mapped_params["maxOutputTokens"] = mapped_params.pop("max_tokens")
|
||||
if "top_k" in mapped_params:
|
||||
mapped_params["topK"] = mapped_params.pop("top_k")
|
||||
if "top_p" in mapped_params:
|
||||
mapped_params["topP"] = mapped_params.pop("top_p")
|
||||
|
||||
unsupported = ["frequency_penalty", "presence_penalty", "repetition_penalty"]
|
||||
for param in unsupported:
|
||||
if param in mapped_params:
|
||||
logger.warning(f"Gemini 原生API不支持参数 '{param}'，已忽略")
|
||||
mapped_params.pop(param)
|
||||
|
||||
elif api_type in ["openai", "deepseek", "zhipu", "general_openai_compat"]:
|
||||
if "repetition_penalty" in mapped_params and api_type == "openai":
|
||||
logger.warning("OpenAI官方API不支持repetition_penalty参数，已忽略")
|
||||
mapped_params.pop("repetition_penalty")
|
||||
|
||||
if "stop" in mapped_params:
|
||||
stop_value = mapped_params["stop"]
|
||||
if isinstance(stop_value, str):
|
||||
mapped_params["stop"] = [stop_value]
|
||||
|
||||
return mapped_params
|
||||
|
||||
|
||||
def create_generation_config_from_kwargs(**kwargs) -> LLMGenerationConfig:
|
||||
"""从关键字参数创建生成配置"""
|
||||
model_fields = getattr(LLMGenerationConfig, "model_fields", {})
|
||||
known_fields = set(model_fields.keys())
|
||||
known_params = {}
|
||||
custom_params = {}
|
||||
|
||||
for key, value in kwargs.items():
|
||||
if key in known_fields:
|
||||
known_params[key] = value
|
||||
else:
|
||||
custom_params[key] = value
|
||||
|
||||
if custom_params:
|
||||
known_params["custom_params"] = custom_params
|
||||
|
||||
return LLMGenerationConfig(**known_params)
|
||||
@@ -0,0 +1,169 @@
|
||||
"""
|
||||
LLM 预设配置
|
||||
|
||||
提供常用的配置预设，特别是针对 Gemini 的高级功能。
|
||||
"""
|
||||
|
||||
from typing import Any
|
||||
|
||||
from .generation import LLMGenerationConfig
|
||||
|
||||
|
||||
class CommonOverrides:
|
||||
"""常用的配置覆盖预设"""
|
||||
|
||||
@staticmethod
|
||||
def creative() -> LLMGenerationConfig:
|
||||
"""创意模式：高温度，鼓励创新"""
|
||||
return LLMGenerationConfig(temperature=0.9, top_p=0.95, frequency_penalty=0.1)
|
||||
|
||||
@staticmethod
|
||||
def precise() -> LLMGenerationConfig:
|
||||
"""精确模式：低温度，确定性输出"""
|
||||
return LLMGenerationConfig(temperature=0.1, top_p=0.9, frequency_penalty=0.0)
|
||||
|
||||
@staticmethod
|
||||
def balanced() -> LLMGenerationConfig:
|
||||
"""平衡模式：中等温度"""
|
||||
return LLMGenerationConfig(temperature=0.5, top_p=0.9, frequency_penalty=0.0)
|
||||
|
||||
@staticmethod
|
||||
def concise(max_tokens: int = 100) -> LLMGenerationConfig:
|
||||
"""简洁模式：限制输出长度"""
|
||||
return LLMGenerationConfig(
|
||||
temperature=0.3,
|
||||
max_tokens=max_tokens,
|
||||
stop=["\n\n", "。", "！", "？"],
|
||||
)
|
||||
|
||||
@staticmethod
|
||||
def detailed(max_tokens: int = 2000) -> LLMGenerationConfig:
|
||||
"""详细模式：鼓励详细输出"""
|
||||
return LLMGenerationConfig(
|
||||
temperature=0.7, max_tokens=max_tokens, frequency_penalty=-0.1
|
||||
)
|
||||
|
||||
@staticmethod
|
||||
def gemini_json() -> LLMGenerationConfig:
|
||||
"""Gemini JSON模式：强制JSON输出"""
|
||||
return LLMGenerationConfig(
|
||||
temperature=0.3, response_mime_type="application/json"
|
||||
)
|
||||
|
||||
@staticmethod
|
||||
def gemini_thinking(budget: float = 0.8) -> LLMGenerationConfig:
|
||||
"""Gemini 思考模式：使用思考预算"""
|
||||
return LLMGenerationConfig(temperature=0.7, thinking_budget=budget)
|
||||
|
||||
@staticmethod
|
||||
def gemini_creative() -> LLMGenerationConfig:
|
||||
"""Gemini 创意模式：高温度创意输出"""
|
||||
return LLMGenerationConfig(temperature=0.9, top_p=0.95)
|
||||
|
||||
@staticmethod
|
||||
def gemini_structured(schema: dict[str, Any]) -> LLMGenerationConfig:
|
||||
"""Gemini 结构化输出：自定义JSON模式"""
|
||||
return LLMGenerationConfig(
|
||||
temperature=0.3,
|
||||
response_mime_type="application/json",
|
||||
response_schema=schema,
|
||||
)
|
||||
|
||||
@staticmethod
|
||||
def gemini_safe() -> LLMGenerationConfig:
|
||||
"""Gemini 安全模式：严格安全设置"""
|
||||
return LLMGenerationConfig(
|
||||
temperature=0.5,
|
||||
safety_settings={
|
||||
"HARM_CATEGORY_HARASSMENT": "BLOCK_MEDIUM_AND_ABOVE",
|
||||
"HARM_CATEGORY_HATE_SPEECH": "BLOCK_MEDIUM_AND_ABOVE",
|
||||
"HARM_CATEGORY_SEXUALLY_EXPLICIT": "BLOCK_MEDIUM_AND_ABOVE",
|
||||
"HARM_CATEGORY_DANGEROUS_CONTENT": "BLOCK_MEDIUM_AND_ABOVE",
|
||||
},
|
||||
)
|
||||
|
||||
@staticmethod
|
||||
def gemini_multimodal() -> LLMGenerationConfig:
|
||||
"""Gemini 多模态模式：优化多模态处理"""
|
||||
return LLMGenerationConfig(temperature=0.6, max_tokens=2048, top_p=0.8)
|
||||
|
||||
@staticmethod
|
||||
def gemini_code_execution() -> LLMGenerationConfig:
|
||||
"""Gemini 代码执行模式：启用代码执行功能"""
|
||||
return LLMGenerationConfig(
|
||||
temperature=0.3,
|
||||
max_tokens=4096,
|
||||
enable_code_execution=True,
|
||||
custom_params={"code_execution_timeout": 30},
|
||||
)
|
||||
|
||||
@staticmethod
|
||||
def gemini_grounding() -> LLMGenerationConfig:
|
||||
"""Gemini 信息来源关联模式：启用Google搜索"""
|
||||
return LLMGenerationConfig(
|
||||
temperature=0.5,
|
||||
max_tokens=4096,
|
||||
enable_grounding=True,
|
||||
custom_params={
|
||||
"grounding_config": {"dynamicRetrievalConfig": {"mode": "MODE_DYNAMIC"}}
|
||||
},
|
||||
)
|
||||
|
||||
@staticmethod
|
||||
def gemini_cached() -> LLMGenerationConfig:
|
||||
"""Gemini 缓存模式：启用响应缓存"""
|
||||
return LLMGenerationConfig(
|
||||
temperature=0.3,
|
||||
max_tokens=2048,
|
||||
enable_caching=True,
|
||||
)
|
||||
|
||||
@staticmethod
|
||||
def gemini_advanced() -> LLMGenerationConfig:
|
||||
"""Gemini 高级模式：启用所有高级功能"""
|
||||
return LLMGenerationConfig(
|
||||
temperature=0.5,
|
||||
max_tokens=4096,
|
||||
enable_code_execution=True,
|
||||
enable_grounding=True,
|
||||
enable_caching=True,
|
||||
custom_params={
|
||||
"code_execution_timeout": 30,
|
||||
"grounding_config": {
|
||||
"dynamicRetrievalConfig": {"mode": "MODE_DYNAMIC"}
|
||||
},
|
||||
},
|
||||
)
|
||||
|
||||
@staticmethod
|
||||
def gemini_research() -> LLMGenerationConfig:
|
||||
"""Gemini 研究模式：思考+搜索+结构化输出"""
|
||||
return LLMGenerationConfig(
|
||||
temperature=0.6,
|
||||
max_tokens=4096,
|
||||
thinking_budget=0.8,
|
||||
enable_grounding=True,
|
||||
response_mime_type="application/json",
|
||||
custom_params={
|
||||
"grounding_config": {"dynamicRetrievalConfig": {"mode": "MODE_DYNAMIC"}}
|
||||
},
|
||||
)
|
||||
|
||||
@staticmethod
|
||||
def gemini_analysis() -> LLMGenerationConfig:
|
||||
"""Gemini 分析模式：深度思考+详细输出"""
|
||||
return LLMGenerationConfig(
|
||||
temperature=0.4,
|
||||
max_tokens=6000,
|
||||
thinking_budget=0.9,
|
||||
top_p=0.8,
|
||||
)
|
||||
|
||||
@staticmethod
|
||||
def gemini_fast_response() -> LLMGenerationConfig:
|
||||
"""Gemini 快速响应模式：低延迟+简洁输出"""
|
||||
return LLMGenerationConfig(
|
||||
temperature=0.3,
|
||||
max_tokens=512,
|
||||
top_p=0.8,
|
||||
)
|
||||
@@ -0,0 +1,328 @@
|
||||
"""
|
||||
LLM 提供商配置管理
|
||||
|
||||
负责注册和管理 AI 服务提供商的配置项。
|
||||
"""
|
||||
|
||||
from typing import Any
|
||||
|
||||
from pydantic import BaseModel, Field
|
||||
|
||||
from zhenxun.configs.config import Config
|
||||
from zhenxun.services.log import logger
|
||||
|
||||
from ..types.models import ModelDetail, ProviderConfig
|
||||
|
||||
AI_CONFIG_GROUP = "AI"
|
||||
PROVIDERS_CONFIG_KEY = "PROVIDERS"
|
||||
|
||||
|
||||
class LLMConfig(BaseModel):
|
||||
"""LLM 服务配置类"""
|
||||
|
||||
default_model_name: str | None = Field(
|
||||
default=None,
|
||||
description="LLM服务全局默认使用的模型名称 (格式: ProviderName/ModelName)",
|
||||
)
|
||||
proxy: str | None = Field(
|
||||
default=None,
|
||||
description="LLM服务请求使用的网络代理，例如 http://127.0.0.1:7890",
|
||||
)
|
||||
timeout: int = Field(default=180, description="LLM服务API请求超时时间（秒）")
|
||||
max_retries_llm: int = Field(
|
||||
default=3, description="LLM服务请求失败时的最大重试次数"
|
||||
)
|
||||
retry_delay_llm: int = Field(
|
||||
default=2, description="LLM服务请求重试的基础延迟时间（秒）"
|
||||
)
|
||||
providers: list[ProviderConfig] = Field(
|
||||
default_factory=list, description="配置多个 AI 服务提供商及其模型信息"
|
||||
)
|
||||
|
||||
def get_provider_by_name(self, name: str) -> ProviderConfig | None:
|
||||
"""根据名称获取提供商配置
|
||||
|
||||
参数:
|
||||
name: 提供商名称
|
||||
|
||||
返回:
|
||||
ProviderConfig | None: 提供商配置，如果未找到则返回 None
|
||||
"""
|
||||
for provider in self.providers:
|
||||
if provider.name == name:
|
||||
return provider
|
||||
return None
|
||||
|
||||
def get_model_by_provider_and_name(
|
||||
self, provider_name: str, model_name: str
|
||||
) -> tuple[ProviderConfig, ModelDetail] | None:
|
||||
"""根据提供商名称和模型名称获取配置
|
||||
|
||||
参数:
|
||||
provider_name: 提供商名称
|
||||
model_name: 模型名称
|
||||
|
||||
返回:
|
||||
tuple[ProviderConfig, ModelDetail] | None: 提供商配置和模型详情的元组，
|
||||
如果未找到则返回 None
|
||||
"""
|
||||
provider = self.get_provider_by_name(provider_name)
|
||||
if not provider:
|
||||
return None
|
||||
|
||||
for model in provider.models:
|
||||
if model.model_name == model_name:
|
||||
return provider, model
|
||||
return None
|
||||
|
||||
def list_available_models(self) -> list[dict[str, Any]]:
|
||||
"""列出所有可用的模型
|
||||
|
||||
返回:
|
||||
list[dict[str, Any]]: 模型信息列表
|
||||
"""
|
||||
models = []
|
||||
for provider in self.providers:
|
||||
for model in provider.models:
|
||||
models.append(
|
||||
{
|
||||
"provider_name": provider.name,
|
||||
"model_name": model.model_name,
|
||||
"full_name": f"{provider.name}/{model.model_name}",
|
||||
"is_available": model.is_available,
|
||||
"is_embedding_model": model.is_embedding_model,
|
||||
"api_type": provider.api_type,
|
||||
}
|
||||
)
|
||||
return models
|
||||
|
||||
def validate_model_name(self, provider_model_name: str) -> bool:
|
||||
"""验证模型名称格式是否正确
|
||||
|
||||
参数:
|
||||
provider_model_name: 格式为 "ProviderName/ModelName" 的字符串
|
||||
|
||||
返回:
|
||||
bool: 是否有效
|
||||
"""
|
||||
if not provider_model_name or "/" not in provider_model_name:
|
||||
return False
|
||||
|
||||
parts = provider_model_name.split("/", 1)
|
||||
if len(parts) != 2:
|
||||
return False
|
||||
|
||||
provider_name, model_name = parts
|
||||
return (
|
||||
self.get_model_by_provider_and_name(provider_name, model_name) is not None
|
||||
)
|
||||
|
||||
|
||||
def get_ai_config():
|
||||
"""获取 AI 配置组"""
|
||||
return Config.get(AI_CONFIG_GROUP)
|
||||
|
||||
|
||||
def get_default_providers() -> list[dict[str, Any]]:
|
||||
"""获取默认的提供商配置
|
||||
|
||||
返回:
|
||||
list[dict[str, Any]]: 默认提供商配置列表
|
||||
"""
|
||||
return [
|
||||
{
|
||||
"name": "DeepSeek",
|
||||
"api_key": "sk-******",
|
||||
"api_base": "https://api.deepseek.com",
|
||||
"api_type": "openai",
|
||||
"models": [
|
||||
{
|
||||
"model_name": "deepseek-chat",
|
||||
"max_tokens": 4096,
|
||||
"temperature": 0.7,
|
||||
},
|
||||
{
|
||||
"model_name": "deepseek-reasoner",
|
||||
},
|
||||
],
|
||||
},
|
||||
{
|
||||
"name": "GLM",
|
||||
"api_key": "",
|
||||
"api_base": "https://open.bigmodel.cn",
|
||||
"api_type": "zhipu",
|
||||
"models": [
|
||||
{"model_name": "glm-4-flash"},
|
||||
{"model_name": "glm-4-plus"},
|
||||
],
|
||||
},
|
||||
{
|
||||
"name": "Gemini",
|
||||
"api_key": [
|
||||
"AIzaSy*****************************",
|
||||
"AIzaSy*****************************",
|
||||
"AIzaSy*****************************",
|
||||
],
|
||||
"api_base": "https://generativelanguage.googleapis.com",
|
||||
"api_type": "gemini",
|
||||
"models": [
|
||||
{"model_name": "gemini-2.0-flash"},
|
||||
{"model_name": "gemini-2.5-flash-preview-05-20"},
|
||||
],
|
||||
},
|
||||
]
|
||||
|
||||
|
||||
def register_llm_configs():
|
||||
"""注册 LLM 服务的配置项"""
|
||||
logger.info("注册 LLM 服务的配置项")
|
||||
|
||||
llm_config = LLMConfig()
|
||||
|
||||
Config.add_plugin_config(
|
||||
AI_CONFIG_GROUP,
|
||||
"default_model_name",
|
||||
llm_config.default_model_name,
|
||||
help="LLM服务全局默认使用的模型名称 (格式: ProviderName/ModelName)",
|
||||
type=str,
|
||||
)
|
||||
Config.add_plugin_config(
|
||||
AI_CONFIG_GROUP,
|
||||
"proxy",
|
||||
llm_config.proxy,
|
||||
help="LLM服务请求使用的网络代理，例如 http://127.0.0.1:7890",
|
||||
type=str,
|
||||
)
|
||||
Config.add_plugin_config(
|
||||
AI_CONFIG_GROUP,
|
||||
"timeout",
|
||||
llm_config.timeout,
|
||||
help="LLM服务API请求超时时间（秒）",
|
||||
type=int,
|
||||
)
|
||||
Config.add_plugin_config(
|
||||
AI_CONFIG_GROUP,
|
||||
"max_retries_llm",
|
||||
llm_config.max_retries_llm,
|
||||
help="LLM服务请求失败时的最大重试次数",
|
||||
type=int,
|
||||
)
|
||||
Config.add_plugin_config(
|
||||
AI_CONFIG_GROUP,
|
||||
"retry_delay_llm",
|
||||
llm_config.retry_delay_llm,
|
||||
help="LLM服务请求重试的基础延迟时间（秒）",
|
||||
type=int,
|
||||
)
|
||||
|
||||
Config.add_plugin_config(
|
||||
AI_CONFIG_GROUP,
|
||||
PROVIDERS_CONFIG_KEY,
|
||||
get_default_providers(),
|
||||
help="配置多个 AI 服务提供商及其模型信息",
|
||||
default_value=[],
|
||||
type=list[ProviderConfig],
|
||||
)
|
||||
|
||||
|
||||
def get_llm_config() -> LLMConfig:
|
||||
"""获取 LLM 配置实例
|
||||
|
||||
返回:
|
||||
LLMConfig: LLM 配置实例
|
||||
"""
|
||||
ai_config = get_ai_config()
|
||||
|
||||
config_data = {
|
||||
"default_model_name": ai_config.get("default_model_name"),
|
||||
"proxy": ai_config.get("proxy"),
|
||||
"timeout": ai_config.get("timeout", 180),
|
||||
"max_retries_llm": ai_config.get("max_retries_llm", 3),
|
||||
"retry_delay_llm": ai_config.get("retry_delay_llm", 2),
|
||||
"providers": ai_config.get(PROVIDERS_CONFIG_KEY, []),
|
||||
}
|
||||
|
||||
return LLMConfig(**config_data)
|
||||
|
||||
|
||||
def validate_llm_config() -> tuple[bool, list[str]]:
|
||||
"""验证 LLM 配置的有效性
|
||||
|
||||
返回:
|
||||
tuple[bool, list[str]]: (是否有效, 错误信息列表)
|
||||
"""
|
||||
errors = []
|
||||
|
||||
try:
|
||||
llm_config = get_llm_config()
|
||||
|
||||
if llm_config.timeout <= 0:
|
||||
errors.append("timeout 必须大于 0")
|
||||
|
||||
if llm_config.max_retries_llm < 0:
|
||||
errors.append("max_retries_llm 不能小于 0")
|
||||
|
||||
if llm_config.retry_delay_llm <= 0:
|
||||
errors.append("retry_delay_llm 必须大于 0")
|
||||
|
||||
if not llm_config.providers:
|
||||
errors.append("至少需要配置一个 AI 服务提供商")
|
||||
else:
|
||||
provider_names = set()
|
||||
for provider in llm_config.providers:
|
||||
if provider.name in provider_names:
|
||||
errors.append(f"提供商名称重复: {provider.name}")
|
||||
provider_names.add(provider.name)
|
||||
|
||||
if not provider.api_key:
|
||||
errors.append(f"提供商 {provider.name} 缺少 API Key")
|
||||
|
||||
if not provider.models:
|
||||
errors.append(f"提供商 {provider.name} 没有配置任何模型")
|
||||
else:
|
||||
model_names = set()
|
||||
for model in provider.models:
|
||||
if model.model_name in model_names:
|
||||
errors.append(
|
||||
f"提供商 {provider.name} 中模型名称重复: "
|
||||
f"{model.model_name}"
|
||||
)
|
||||
model_names.add(model.model_name)
|
||||
|
||||
if llm_config.default_model_name:
|
||||
if not llm_config.validate_model_name(llm_config.default_model_name):
|
||||
errors.append(
|
||||
f"默认模型 {llm_config.default_model_name} 在配置中不存在"
|
||||
)
|
||||
|
||||
except Exception as e:
|
||||
errors.append(f"配置解析失败: {e!s}")
|
||||
|
||||
return len(errors) == 0, errors
|
||||
|
||||
|
||||
def set_default_model(provider_model_name: str | None) -> bool:
|
||||
"""设置默认模型
|
||||
|
||||
参数:
|
||||
provider_model_name: 模型名称，格式为 "ProviderName/ModelName"，None 表示清除
|
||||
|
||||
返回:
|
||||
bool: 是否设置成功
|
||||
"""
|
||||
if provider_model_name:
|
||||
llm_config = get_llm_config()
|
||||
if not llm_config.validate_model_name(provider_model_name):
|
||||
logger.error(f"模型 {provider_model_name} 在配置中不存在")
|
||||
return False
|
||||
|
||||
Config.set_config(
|
||||
AI_CONFIG_GROUP, "default_model_name", provider_model_name, auto_save=True
|
||||
)
|
||||
|
||||
if provider_model_name:
|
||||
logger.info(f"默认模型已设置为: {provider_model_name}")
|
||||
else:
|
||||
logger.info("默认模型已清除")
|
||||
|
||||
return True
|
||||
Reference in New Issue
Block a user