Update app.py

app.py

@@ -25,7 +25,9 @@ model = LlamaForCausalLM.from_pretrained(
     model_name,
     quantization_config=quant_config
 )
-model = PeftModel.from_pretrained(model, lora_name)
+
+config = PeftConfig.from_pretrained(lora_name, device="cpu")
+model = PeftModel.from_pretrained(model, lora_name, config=config)
 
 model = model.to("cuda")