Update app.py

app.py

@@ -6,6 +6,9 @@ llm = Llama.from_pretrained(
     repo_id="bartowski/Phi-3.5-mini-instruct-GGUF",
     filename="Phi-3.5-mini-instruct-Q4_K_M.gguf",
     numa=True,
+    use_mmap=False,
+    use_mlock=True,
+    seed=-1,
     # flash_attn=True,
     # n_gpu_layers=-1,
     n_batch=1024,