Update app.py

app.py

@@ -18,12 +18,12 @@ def generate_response(prompt):
     # 모델 및 토크나이저 로딩은 함수 내부에서 수행
     # tokenizer = AutoTokenizer.from_pretrained("dasomaru/gemma-3-4bit-it-demo")
     # model = AutoModelForCausalLM.from_pretrained("dasomaru/gemma-3-4bit-it-demo")
-    # tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
-    # model = AutoModelForCausalLM.from_pretrained(
-    #     model_name,
-    #     torch_dtype=torch.float16,    
-    #     trust_remote_code=True,
-    # )
+    tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
+    model = AutoModelForCausalLM.from_pretrained(
+        model_name,
+        torch_dtype=torch.float16,    
+        trust_remote_code=True,
+    )
     model.to("cuda")    
 
     # inputs = tokenizer(prompt, return_tensors="pt").to(model.device)