Update app.py

app.py

@@ -23,7 +23,8 @@ quant_config = BitsAndBytesConfig(
 tokenizer = LlamaTokenizer.from_pretrained(model_name)
 model = LlamaForCausalLM.from_pretrained(
     model_name,
-    quantization_config=quant_config
+    device_map="cpu"
+    #quantization_config=quant_config
 )
 
 model = PeftModel.from_pretrained(model, lora_name, hf_device_map="cpu")