new code

src/pdfchatbot.py

@@ -56,7 +56,7 @@ class PDFChatBot:
         print("Vector store created")
     @spaces.GPU
     def load_tokenizer(self):
-        self.tokenizer = AutoModelForCausalLM.from_pretrained("meta-llama/Meta-Llama-3-8B-Instruct")
+        self.tokenizer = AutoTokenizer.from_pretrained("meta-llama/Meta-Llama-3-8B-Instruct")
 
     @spaces.GPU
     def create_organic_pipeline(self):
@@ -66,7 +66,6 @@ class PDFChatBot:
             model_kwargs={"torch_dtype": torch.bfloat16},
             device="cuda",
         )
-        self.load_tokenizer()
         print("Model pipeline loaded")
 
     def get_organic_context(self, query):
@@ -78,7 +77,6 @@ class PDFChatBot:
     @spaces.GPU
     def create_organic_response(self, history, query):
         self.get_organic_context(query)
-        tokenizer = AutoTokenizer.from_pretrained("meta-llama/Meta-Llama-3-8B-Instruct")
         """
         pipe = pipeline(
             "text-generation",
@@ -92,7 +90,7 @@ class PDFChatBot:
             {"role": "user", "content": query},
         ]
 
-        prompt = tokenizer.apply_chat_template(
+        prompt = self.pipe.tokenizer.apply_chat_template(
             messages,
             tokenize=False,
             add_generation_prompt=True