Update app.py

app.py

@@ -25,7 +25,7 @@ if not torch.cuda.is_available():
 
 if torch.cuda.is_available():
     model_id = "NousResearch/Hermes-2-Pro-Mistral-7B"
-    model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto", load_in_4bit=True)
+    model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto")
     tokenizer = AutoTokenizer.from_pretrained(model_id)
     tokenizer.use_default_system_prompt = False