fix(chatbot): reduce timeout settings and enhance error handling

Author: alienx5499

SortVision/.env.example

@@ -43,10 +43,10 @@ NVIDIA_API_KEY=your_nvidia_api_key_here
 NVIDIA_BASE_URL=https://integrate.api.nvidia.com/v1
 NVIDIA_MODEL=moonshotai/kimi-k2-instruct
 NVIDIA_MODEL_FALLBACKS=
-NVIDIA_TIMEOUT_MS=15000
+NVIDIA_TIMEOUT_MS=7000
 NVIDIA_TEMPERATURE=0.6
 NVIDIA_TOP_P=0.9
-NVIDIA_MAX_TOKENS=1024
+NVIDIA_MAX_TOKENS=512
 
 # Chat moderation and CORS controls
 NEXT_PUBLIC_CHAT_ABUSE_THRESHOLD=3

SortVision/src/app/api/chatbot/route.js

@@ -9,10 +9,10 @@ const MODEL_FALLBACKS = (process.env.NVIDIA_MODEL_FALLBACKS || '')
   .map(model => model.trim())
   .filter(Boolean);
 
-const REQUEST_TIMEOUT_MS = Number(process.env.NVIDIA_TIMEOUT_MS || 15000);
+const REQUEST_TIMEOUT_MS = Number(process.env.NVIDIA_TIMEOUT_MS || 7000);
 const TEMPERATURE = Number(process.env.NVIDIA_TEMPERATURE || 0.6);
 const TOP_P = Number(process.env.NVIDIA_TOP_P || 0.9);
-const MAX_TOKENS = Number(process.env.NVIDIA_MAX_TOKENS || 1024);
+const MAX_TOKENS = Math.min(Number(process.env.NVIDIA_MAX_TOKENS || 512), 512);
 const ABUSE_THRESHOLD = Number(process.env.CHAT_ABUSE_THRESHOLD || 3);
 const ABUSE_WINDOW_MS = Number(
   process.env.CHAT_ABUSE_WINDOW_MS || 10 * 60 * 1000
@@ -256,8 +256,12 @@ export async function POST(req) {
     const modelsToTry = getModelsToTry();
     let completion = null;
     let lastError = null;
+    const requestStart = Date.now();
 
     for (const model of modelsToTry) {
+      if (Date.now() - requestStart > REQUEST_TIMEOUT_MS - 500) {
+        break;
+      }
       try {
         completion = await client.chat.completions.create({
           model,
@@ -276,6 +280,14 @@ export async function POST(req) {
             error?.message || error
           );
         }
+        const isTimeoutError =
+          error?.name?.toLowerCase?.().includes('timeout') ||
+          String(error?.message || '')
+            .toLowerCase()
+            .includes('timeout');
+        if (isTimeoutError) {
+          break;
+        }
       }
     }
 

SortVision/src/components/chatbot/assistantEngine/aiClient.js

@@ -58,7 +58,7 @@ Current sorting context:
 
     try {
       const controller = new AbortController();
-      const timeoutId = setTimeout(() => controller.abort(), 10000);
+      const timeoutId = setTimeout(() => controller.abort(), 8500);
 
       const res = await fetch(CHAT_API_ENDPOINT, {
         method: 'POST',
@@ -114,6 +114,9 @@ Current sorting context:
       if (err.message.includes('API Error: 500')) {
         throw new Error('SERVER_ERROR', { cause: err });
       }
+      if (err.message.includes('API Error: 504')) {
+        throw new Error('TIMEOUT_ERROR', { cause: err });
+      }
       if (err.message.includes('API Error: 429')) {
         throw new Error('RATE_LIMIT', { cause: err });
       }