remove cache

app.py

@@ -38,8 +38,8 @@ def create_knowledge_base(chunks):
 # Hugging Face 모델 로드
 def load_model():
     model_name = "halyn/gemma2-2b-it-finetuned-paperqa"  # 텍스트 생성 모델 사용
-    tokenizer = AutoTokenizer.from_pretrained(model_name, clean_up_tokenization_spaces=False)
-    model = AutoModelForCausalLM.from_pretrained(model_name)
+    tokenizer = AutoTokenizer.from_pretrained(model_name, cache_dir=None, clean_up_tokenization_spaces=False)
+    model = AutoModelForCausalLM.from_pretrained(model_name, cache_dir=None)
     return pipeline("text-generation", model=model, tokenizer=tokenizer, max_new_tokens=150, temperature=0.1)
 
 # QA 체인 설정