feat: 更新模型配置和选择逻辑，修改为通过 model_spec 统一指定模型格式

Author: xerrors

docs/intro/model-config.md

@@ -21,6 +21,16 @@
 
 <<< @/../.env.template#model_provider{bash 2}
 
+### 默认对话模型格式
+
+系统的默认对话模型通过配置项 `default_model` 指定，格式统一为 `模型提供商/模型名称`，例如：
+
+```yaml
+default_model: siliconflow/deepseek-ai/DeepSeek-V3.2-Exp
+```
+
+在 Web 界面中选择模型时也会自动按照这一格式保存，无需手动拆分提供商和模型名称。
+
 
 ::: tip 免费获取 API Key
 [硅基流动](https://cloud.siliconflow.cn/i/Eo5yTHGJ) 注册即送 14 元额度，支持多种开源模型。

server/routers/chat_router.py

@@ -85,7 +85,11 @@ async def set_default_agent(request_data: dict = Body(...), current_user=Depends
 async def call(query: str = Body(...), meta: dict = Body(None), current_user: User = Depends(get_required_user)):
     """调用模型进行简单问答（需要登录）"""
     meta = meta or {}
-    model = select_model(model_provider=meta.get("model_provider"), model_name=meta.get("model_name"))
+    model = select_model(
+        model_provider=meta.get("model_provider"),
+        model_name=meta.get("model_name"),
+        model_spec=meta.get("model_spec") or meta.get("model"),
+    )
 
     async def call_async(query):
         loop = asyncio.get_event_loop()

src/config/app.py

@@ -58,8 +58,11 @@ def __init__(self):
         # 模型配置
         ## 注意这里是模型名，而不是具体的模型路径，默认使用 HuggingFace 的路径
         ## 如果需要自定义本地模型路径，则在 .env 中配置 MODEL_DIR
-        self.add_item("model_provider", default="siliconflow", des="模型提供商", choices=list(self.model_names.keys()))
-        self.add_item("model_name", default="zai-org/GLM-4.5", des="模型名称")
+        self.add_item(
+            "default_model",
+            default=self._get_default_chat_model_spec(),
+            des="默认对话模型",
+        )
         self.add_item(
             "fast_model",
             default="siliconflow/THUDM/GLM-4-9B-0414",
@@ -81,6 +84,9 @@ def __init__(self):
         ### <<< 默认配置结束
 
         self.load()
+        # 清理已废弃的配置项
+        self.pop("model_provider", None)
+        self.pop("model_name", None)
         self.handle_self()
 
     def add_item(self, key, default, des=None, choices=None):
@@ -137,6 +143,22 @@ def _save_models_to_file(self):
         with open(self._models_config_path, "w", encoding="utf-8") as f:
             yaml.safe_dump(models_payload, f, indent=2, allow_unicode=True, sort_keys=False)
 
+    def _get_default_chat_model_spec(self):
+        """选择一个默认的聊天模型，优先使用 siliconflow 的默认模型"""
+        preferred_provider = "siliconflow"
+        provider_info = (self.model_names or {}).get(preferred_provider)
+        if provider_info:
+            default_model = provider_info.get("default")
+            if default_model:
+                return f"{preferred_provider}/{default_model}"
+
+        for provider, info in (self.model_names or {}).items():
+            default_model = info.get("default")
+            if default_model:
+                return f"{provider}/{default_model}"
+
+        return ""
+
     def handle_self(self):
         """
         处理配置

src/knowledge/implementations/lightrag.py

@@ -174,16 +174,19 @@ def _get_llm_func(self, llm_info: dict):
         from src.models import select_model
 
         # 如果用户选择了LLM，使用用户选择的；否则使用环境变量默认值
-        if llm_info and llm_info.get("provider") and llm_info.get("model_name"):
-            provider = llm_info["provider"]
-            model_name = llm_info["model_name"]
+        if llm_info and llm_info.get("model_spec"):
+            model_spec = llm_info["model_spec"]
+            logger.info(f"Using user-selected LLM spec: {model_spec}")
+        elif llm_info and llm_info.get("provider") and llm_info.get("model_name"):
+            model_spec = f"{llm_info['provider']}/{llm_info['model_name']}"
             logger.info(f"Using user-selected LLM: {provider}/{model_name}")
         else:
             provider = LIGHTRAG_LLM_PROVIDER
             model_name = LIGHTRAG_LLM_NAME
+            model_spec = f"{provider}/{model_name}"
             logger.info(f"Using default LLM from environment: {provider}/{model_name}")
 
-        model = select_model(provider, model_name)
+        model = select_model(model_spec=model_spec)
 
         async def llm_model_func(prompt, system_prompt=None, history_messages=[], **kwargs):
             return await openai_complete_if_cache(

src/models/chat.py

@@ -8,6 +8,21 @@
 from src.utils import logger
 
 
+def split_model_spec(model_spec, sep="/"):
+    """
+    将 provider/model 形式的字符串拆分为 (provider, model)
+    """
+    if not model_spec or not isinstance(model_spec, str):
+        return "", ""
+    if not sep:
+        return model_spec, ""
+    try:
+        provider, model_name = model_spec.split(sep, 1)
+        return provider, model_name
+    except ValueError:
+        return model_spec, ""
+
+
 class OpenAIBase:
     def __init__(self, api_key, base_url, model_name, **kwargs):
         self.api_key = api_key
@@ -85,12 +100,26 @@ def __init__(self, content):
         self.is_full = False
 
 
-def select_model(model_provider, model_name=None):
+def select_model(model_provider=None, model_name=None, model_spec=None):
     """根据模型提供者选择模型"""
-    assert model_provider is not None, "Model provider not specified"
+    if model_spec:
+        spec_provider, spec_model_name = split_model_spec(model_spec)
+        model_provider = model_provider or spec_provider
+        model_name = model_name or spec_model_name
+
+    if model_provider is None or not model_name:
+        default_provider, default_model = split_model_spec(getattr(config, "default_model", ""))
+        model_provider = model_provider or default_provider
+        model_name = model_name or default_model
+
+    assert model_provider, "Model provider not specified"
+
     model_info = config.model_names.get(model_provider, {})
     model_name = model_name or model_info.get("default", "")
 
+    if not model_name:
+        raise ValueError(f"Model name not specified for provider {model_provider}")
+
     logger.info(f"Selecting model from `{model_provider}` with `{model_name}`")
 
     if model_provider == "openai":

test/bruteforce_simulation.py

@@ -23,9 +23,7 @@ def parse_args() -> argparse.Namespace:
     parser = argparse.ArgumentParser(description="Simulate brute-force login attempts.")
     parser.add_argument("--base-url", default=os.getenv("TEST_BASE_URL", "http://localhost:5050"), help="API base URL")
     parser.add_argument("--username", default=os.getenv("TEST_USERNAME", "admin"), help="Login identifier to attack")
-    parser.add_argument(
-        "--attempts", type=int, default=20, help="Total number of attempts to issue (default: 20)"
-    )
+    parser.add_argument("--attempts", type=int, default=20, help="Total number of attempts to issue (default: 20)")
     parser.add_argument(
         "--concurrency",
         type=int,
@@ -68,10 +66,13 @@ async def attempt_login(
         started = time.perf_counter()
         response = await client.post("/api/auth/token", data=payload)
         elapsed = time.perf_counter() - started
-        detail = response.json().get("detail") if response.headers.get("content-type", "").startswith("application/json") else response.text
+        detail = (
+            response.json().get("detail")
+            if response.headers.get("content-type", "").startswith("application/json")
+            else response.text
+        )
         print(
-            f"[{attempt_no:02d}] {response.status_code} in {elapsed*1000:.1f} ms "
-            f"(pwd={password!r}) detail={detail!r}"
+            f"[{attempt_no:02d}] {response.status_code} in {elapsed * 1000:.1f} ms (pwd={password!r}) detail={detail!r}"
         )
         return response.status_code, elapsed
 
@@ -86,9 +87,7 @@ async def run_simulation(args: argparse.Namespace) -> int:
         tasks = []
         for attempt_no in range(1, args.attempts + 1):
             tasks.append(
-                asyncio.create_task(
-                    attempt_login(client, semaphore, attempt_no, args.username, args.password)
-                )
+                asyncio.create_task(attempt_login(client, semaphore, attempt_no, args.username, args.password))
             )
             if args.delay:
                 await asyncio.sleep(args.delay)

web/src/components/AgentConfigSidebar.vue

@@ -56,8 +56,7 @@
                 <div v-if="value.template_metadata.kind === 'llm'" class="model-selector">
                   <ModelSelectorComponent
                     @select-model="handleModelChange"
-                    :model_name="agentConfig[key] ? agentConfig[key].split('/').slice(1).join('/') : ''"
-                    :model_provider="agentConfig[key] ? agentConfig[key].split('/')[0] : ''"
+                    :model_spec="agentConfig[key] || ''"
                   />
                 </div>
 
@@ -367,9 +366,10 @@ const getPlaceholder = (key, value) => {
   return `（默认: ${value.default}）`;
 };
 
-const handleModelChange = (data) => {
+const handleModelChange = (spec) => {
+  if (typeof spec !== 'string' || !spec) return;
   agentStore.updateAgentConfig({
-    model: `${data.provider}/${data.name}`
+    model: spec
   });
 };
 

web/src/components/ModelSelectorComponent.vue

@@ -3,10 +3,10 @@
     <div class="model-select" @click.prevent>
       <div class="model-select-content">
         <div class="model-info">
-          <a-tooltip :title="model_name" placement="right">
-            <span class="model-text text"> {{ model_name }} </span>
+          <a-tooltip :title="displayModelText" placement="right">
+            <span class="model-text text"> {{ displayModelText }} </span>
           </a-tooltip>
-          <span class="model-provider">{{ model_provider }}</span>
+          <span class="model-provider">{{ displayModelProvider }}</span>
         </div>
         <div class="model-status-controls">
           <span
@@ -48,13 +48,17 @@ import { useConfigStore } from '@/stores/config'
 import { chatModelApi } from '@/apis/system_api'
 
 const props = defineProps({
-  model_name: {
+  model_spec: {
     type: String,
     default: ''
   },
-  model_provider: {
+  sep: {
     type: String,
-    default: ''
+    default: '/'
+  },
+  placeholder: {
+    type: String,
+    default: '请选择模型'
   }
 });
 
@@ -76,25 +80,48 @@ const modelKeys = computed(() => {
   return Object.keys(modelStatus.value || {}).filter(key => modelStatus.value?.[key])
 })
 
+const resolvedSep = computed(() => props.sep || '/')
+
+const resolvedModel = computed(() => {
+  const spec = props.model_spec || ''
+  const sep = resolvedSep.value
+  if (spec && sep) {
+    const index = spec.indexOf(sep)
+    if (index !== -1) {
+      const provider = spec.slice(0, index)
+      const name = spec.slice(index + sep.length)
+      if (provider && name) {
+        return { provider, name }
+      }
+    }
+  }
+  return { provider: '', name: '' }
+})
+
+const displayModelProvider = computed(() => resolvedModel.value.provider || '')
+const displayModelName = computed(() => resolvedModel.value.name || '')
+const displayModelText = computed(() => displayModelName.value || props.placeholder)
+
 // 当前模型状态
 const currentModelStatus = computed(() => {
   return state.currentModelStatus
 })
 
 // 检查当前模型状态
 const checkCurrentModelStatus = async () => {
-  if (!props.model_provider || !props.model_name) return
+  const { provider, name } = resolvedModel.value
+  if (!provider || !name) return
 
   try {
     state.checkingStatus = true
-    const response = await chatModelApi.getModelStatus(props.model_provider, props.model_name)
+    const response = await chatModelApi.getModelStatus(provider, name)
     if (response.status) {
       state.currentModelStatus = response.status
     } else {
       state.currentModelStatus = null
     }
   } catch (error) {
-    console.error(`检查当前模型 ${props.model_provider}/${props.model_name} 状态失败:`, error)
+    console.error(`检查当前模型 ${provider}/${name} 状态失败:`, error)
     state.currentModelStatus = { status: 'error', message: error.message }
   } finally {
     state.checkingStatus = false
@@ -128,24 +155,27 @@ const getCurrentModelStatusTooltip = () => {
 
 // 选择模型的方法
 const handleSelectModel = async (provider, name) => {
-  emit('select-model', { provider, name })
+  const sep = resolvedSep.value || '/'
+  const separator = sep || '/'
+  const spec = `${provider}${separator}${name}`
+  emit('select-model', spec)
 }
 
 </script>
 
 <style lang="less" scoped>
 // 变量定义
-@status-success: #52c41a;
-@status-error: #ff4d4f;
-@status-warning: #faad14;
-@status-default: #999;
+@status-success: var(--color-success);
+@status-error: var(--color-error);
+@status-warning: var(--chart-warning);
+@status-default: var(--gray-500);
 @border-radius: 8px;
 @scrollbar-width: 6px;
 @status-indicator-padding: 2px 4px;
 @status-check-button-padding: 0 4px;
 @status-check-button-font-size: 12px;
 @status-indicator-font-size: 11px;
-@model-provider-color: #aaa;
+@model-provider-color: var(--gray-500);
 
 // 主选择器样式
 .model-select {
@@ -156,11 +186,12 @@ const handleSelectModel = async (provider, name) => {
   cursor: pointer;
   border: 1px solid var(--gray-200);
   border-radius: @border-radius;
-  background-color: white;
+  background-color: var(--gray-0);
   min-width: 0;
   display: flex;
   align-items: center;
   gap: 0.5rem;
+  font-size: 13px;
 
   // 修饰符类
   &.borderless {
@@ -188,7 +219,7 @@ const handleSelectModel = async (provider, name) => {
       .model-text {
         overflow: hidden;
         text-overflow: ellipsis;
-        color: #000;
+        color: var(--gray-1000);
         white-space: nowrap;
       }
 
@@ -270,4 +301,4 @@ const handleSelectModel = async (provider, name) => {
     overflow-y: auto;
   }
 }
-</style>
+</style>