Update handler.py

handler.py

@@ -14,10 +14,10 @@ device = "cuda" if torch.cuda.is_available() else "cpu"
 class EndpointHandler():
     def __init__(self, path=""):
         config = PeftConfig.from_pretrained(path)
-        model = AutoModelForCausalLM.from_pretrained(config.base_model_name_or_path, device_map={"":0}, trust_remote_code=True, torch_dtype=torch.float16)
-        self.tokenizer = AutoTokenizer.from_pretrained(config.base_model_name_or_path)
+        model = AutoModelForCausalLM.from_pretrained("vilsonrodrigues/falcon-7b-instruct-sharded", device_map={"":0}, trust_remote_code=True, torch_dtype=torch.float16)
+        self.tokenizer = AutoTokenizer.from_pretrained("vilsonrodrigues/falcon-7b-instruct-sharded")
         # Load the Lora model
-        self.model = PeftModel.from_pretrained(model, path)
+        self.model = PeftModel.from_pretrained(model, "jscore2023/falcon-7b-3")
 
     def __call__(self, data: Dict[str, Any]) -> Dict[str, Any]:
         """