Update app.py

app.py

@@ -23,7 +23,7 @@ try:
         tokenizer=tokenizer,
         torch_dtype=torch.float16,
         num_return_sequences=1,
-        do_sample=True,
+        do_sample=False,
         truncation=True,
         max_new_tokens=128
     )
@@ -44,7 +44,7 @@ def respond(message, history, system_message, max_tokens, temperature, top_p):
             max_new_tokens=max_tokens,
             temperature=temperature,
             top_p=top_p,
-            do_sample=True,
+            do_sample=False,
             pad_token_id=tokenizer.eos_token_id
         )[0]["generated_text"]