updated temperature and fixed filter func

app.py

@@ -41,13 +41,13 @@ def predict (inp_text):
           return_tensors = "pt",
       ).to("cuda")
       model.generation_config.pad_token_id = tokenizer.pad_token_id
-      outputs = model.generate(input_ids = inputs, use_cache = True ,temperature = 0.1,max_new_tokens = 512)
+      outputs = model.generate(input_ids = inputs, use_cache = True ,temperature = 0.01,max_new_tokens = 1024)
       result = tokenizer.batch_decode(outputs)
       # print(result)
       return filter_user_assistant_msgs(result[0])
 
 def filter_user_assistant_msgs(text):
-  msg_pattern = r".*Response:\n(.*?)<\|im_end\|>"
+  msg_pattern = r".*Response:\n(.*?)<\|eot_id\|>"
   match = re.match(msg_pattern, text, re.DOTALL)
   if match:
     message = match.group(1).strip()