refactor: make stream smoothing parameters configurable and add models cache invalidation

- Move stream smoothing parameters (chunk_size, delay_ms) to database config
- Remove hardcoded stream smoothing constants from StreamProcessor
- Simplify dynamic delay calculation by using config values directly
- Add invalidate_models_list_cache() function to clear /v1/models endpoint cache
- Call cache invalidation on model create, update, delete, and bulk operations
- Update admin UI to allow runtime configuration of smoothing parameters
- Improve model listing freshness when models are modified

frontend/src/views/admin/SystemSettings.vue

@@ -470,20 +470,68 @@
         title="流式输出"
         description="配置流式响应的输出效果"
       >
-        <div class="flex items-center space-x-2">
-          <Checkbox
-            id="stream-smoothing-enabled"
-            v-model:checked="systemConfig.stream_smoothing_enabled"
-          />
+        <div class="grid grid-cols-1 md:grid-cols-2 gap-6">
+          <div class="md:col-span-2">
+            <div class="flex items-center space-x-2">
+              <Checkbox
+                id="stream-smoothing-enabled"
+                v-model:checked="systemConfig.stream_smoothing_enabled"
+              />
+              <div>
+                <Label
+                  for="stream-smoothing-enabled"
+                  class="cursor-pointer"
+                >
+                  启用平滑输出
+                </Label>
+                <p class="text-xs text-muted-foreground">
+                  将上游返回的大块内容拆分成小块，模拟打字效果
+                </p>
+              </div>
+            </div>
+          </div>
+
           <div>
             <Label
-              for="stream-smoothing-enabled"
-              class="cursor-pointer"
+              for="stream-smoothing-chunk-size"
+              class="block text-sm font-medium"
             >
-              启用平滑输出
+              每块字符数
             </Label>
-            <p class="text-xs text-muted-foreground">
-              自动根据文本长度调整输出速度：短文本逐字符输出（打字感更强），长文本按块输出（避免卡顿）
+            <Input
+              id="stream-smoothing-chunk-size"
+              v-model.number="systemConfig.stream_smoothing_chunk_size"
+              type="number"
+              min="1"
+              max="100"
+              placeholder="20"
+              class="mt-1"
+              :disabled="!systemConfig.stream_smoothing_enabled"
+            />
+            <p class="mt-1 text-xs text-muted-foreground">
+              每次输出的字符数量（1-100）
+            </p>
+          </div>
+
+          <div>
+            <Label
+              for="stream-smoothing-delay-ms"
+              class="block text-sm font-medium"
+            >
+              输出间隔 (毫秒)
+            </Label>
+            <Input
+              id="stream-smoothing-delay-ms"
+              v-model.number="systemConfig.stream_smoothing_delay_ms"
+              type="number"
+              min="1"
+              max="100"
+              placeholder="8"
+              class="mt-1"
+              :disabled="!systemConfig.stream_smoothing_enabled"
+            />
+            <p class="mt-1 text-xs text-muted-foreground">
+              每块之间的延迟毫秒数（1-100）
             </p>
           </div>
         </div>
@@ -838,6 +886,8 @@ interface SystemConfig {
   audit_log_retention_days: number
   // 流式输出
   stream_smoothing_enabled: boolean
+  stream_smoothing_chunk_size: number
+  stream_smoothing_delay_ms: number
 }
 
 const loading = ref(false)
@@ -889,6 +939,8 @@ const systemConfig = ref<SystemConfig>({
   audit_log_retention_days: 30,
   // 流式输出
   stream_smoothing_enabled: false,
+  stream_smoothing_chunk_size: 20,
+  stream_smoothing_delay_ms: 8,
 })
 
 // 计算属性：KB 和 字节 之间的转换
@@ -947,6 +999,8 @@ async function loadSystemConfig() {
       'audit_log_retention_days',
       // 流式输出
       'stream_smoothing_enabled',
+      'stream_smoothing_chunk_size',
+      'stream_smoothing_delay_ms',
     ]
 
     for (const key of configs) {
@@ -1060,6 +1114,16 @@ async function saveSystemConfig() {
         value: systemConfig.value.stream_smoothing_enabled,
         description: '是否启用流式平滑输出'
       },
+      {
+        key: 'stream_smoothing_chunk_size',
+        value: systemConfig.value.stream_smoothing_chunk_size,
+        description: '流式平滑输出每个小块的字符数'
+      },
+      {
+        key: 'stream_smoothing_delay_ms',
+        value: systemConfig.value.stream_smoothing_delay_ms,
+        description: '流式平滑输出每个小块之间的延迟毫秒数'
+      },
     ]
 
     const promises = configItems.map(item =>