Update app.py

app.py

@@ -28,7 +28,7 @@ server = subprocess.Popen([
     "-ngl", "0",
     "--host", "0.0.0.0",
     "--port", "8080",
-    "--no-mmap",        # avoids page-fault stalls on cold reads
+            # avoids page-fault stalls on cold reads
 ])
 
 print("Starting llama-server, waiting 8 seconds...")
@@ -116,7 +116,7 @@ def chat(user_fact, history):
             {"role": "user",   "content": f"الفعل: {user_fact}"}
         ],
         "stream": True,
-        "temperature": 0.1,   # low = more deterministic / accurate
+        "temperature": 0.9,   # low = more deterministic / accurate
         "top_p": 0.9,
         "max_tokens": 512,    # legal output is short; cap prevents runaway gen
         "repeat_penalty": 1.1,