Update handler.py

handler.py

@@ -8,7 +8,7 @@ class EndpointHandler:
     def __init__(self, model_dir):
         self.model_id = "zesquirrelnator/moondream2-FinetuneV3.1"
         self.model = AutoModelForCausalLM.from_pretrained(self.model_id, trust_remote_code=True)
-        tokenizer = AutoTokenizer.from_pretrained("vikhyatk/moondream2", revision="2024-05-20") # Tokenizer is still from the original repository
+        self.tokenizer = AutoTokenizer.from_pretrained("vikhyatk/moondream2", revision="2024-05-20") # Tokenizer is still from the original repository
         
         # Check if CUDA (GPU support) is available and then set the device to GPU or CPU
         self.device = torch.device("cuda" if torch.cuda.is_available() else "cpu")