Update app.py

app.py

@@ -24,8 +24,10 @@ llm = Llama(
 generation_kwargs = {
     "max_tokens":1000,
     "stop":["</s>"],
+    "temperature":0.2,
     "echo":False, # Echo the prompt in the output
-    "top_k":1 # This is essentially greedy decoding, since the model will always return the highest-probability token. Set this value > 1 for sampling decoding
+    "top_k":20,
+    "top_p":0.7
 }