Update app.py

app.py

@@ -21,7 +21,7 @@ tokenizer = AutoTokenizer.from_pretrained(
 
 ft_model.eval()
 def greet(eval_prompt):
-  model_input = tokenizer(eval_prompt, return_tensors="pt").to("cuda")
+  model_input = tokenizer(eval_prompt, return_tensors="pt").to("cpu")
   with torch.no_grad():
       generation = ft_model.generate(**model_input, max_new_tokens = 150)
       result = tokenizer.decode(generation[0], skip_special_tokens=True)