Update server.py

server.py

@@ -44,7 +44,7 @@ logger.info(f"✓ NVIDIA_API_KEY loaded (length: {len(NVIDIA_API_KEY)})")
 
 # Model configurations
 ROX_CORE_MODEL = "minimaxai/minimax-m2.5"
-ROX_TURBO_MODEL = "deepseek-ai/deepseek-r1-distill-qwen-32b"
+ROX_TURBO_MODEL = "meta/llama-3.1-8b-instruct"  # Changed to a more reliable model
 ROX_CODER_MODEL = "qwen/qwen3.5-397b-a17b"
 ROX_TURBO_45_MODEL = "deepseek-ai/deepseek-v3.1"
 ROX_ULTRA_MODEL = "deepseek-ai/deepseek-v3.2"
@@ -154,7 +154,7 @@ def root():
             "rox_turbo": {
                 "endpoint": "/turbo",
                 "description": "Rox 2.1 Turbo - Fast and efficient",
-                "model": "deepseek-ai/deepseek-r1-distill-qwen-32b",
+                "model": "meta/llama-3.1-8b-instruct",
                 "best_for": "Quick responses and efficient processing"
             },
             "rox_coder": {
@@ -287,16 +287,18 @@ def turbo(req: ChatRequest):
         completion = client.chat.completions.create(
             model=ROX_TURBO_MODEL,
             messages=messages,
-            temperature=req.temperature if req.temperature != 1.0 else 0.6,
-            top_p=req.top_p if req.top_p != 1.0 else 0.7,
+            temperature=req.temperature if req.temperature != 1.0 else 0.7,
+            top_p=req.top_p if req.top_p != 1.0 else 0.9,
             max_tokens=req.max_tokens,
-            stream=False,
+            stream=False
         )
     except Exception as e:
         logger.exception("Error while calling Rox 2.1 Turbo for /turbo")
+        # Log the actual error for debugging
+        logger.error(f"Turbo model error details: {str(e)}")
         raise HTTPException(
             status_code=500,
-            detail="Internal server error while calling Rox 2.1 Turbo.",
+            detail=f"Internal server error while calling Rox 2.1 Turbo: {str(e)}",
         ) from e
 
     try: