change temperature from 0.7 to 0.3 in inference.py

inference.py

@@ -252,7 +252,7 @@ class Chat:
             device,
             num_gpus=1,
             load_8bit=False,
-            temperature=0.7,
+            temperature=0.3, #more conservative
             max_new_tokens=512,
             lora_path=None,
     ):