peft to cuda

app.py

@@ -18,7 +18,7 @@ quant_config = BitsAndBytesConfig(
 model = MistralForCausalLM.from_pretrained("mistralai/Mistral-7B-v0.1", torch_dtype=float16).to("cuda")
     #quantization_config=quant_config,)#.to("cuda")
     #device_map="auto",)
-model = PeftModel.from_pretrained(model, "lang-uk/dragoman")
+model = PeftModel.from_pretrained(model, "lang-uk/dragoman").to("cuda")
 tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-v0.1", use_fast=False, add_bos_token=False)
 
 @spaces.GPU(duration=20)