Update app.py

app.py

@@ -40,7 +40,7 @@ def generate_response(instruction, input_text, max_new_tokens):
                 input_text,   # input
                 ""            # output - leave this blank for generation!
             )
-        ], return_tensors="pt").to("cuda")
+        ], return_tensors="pt").to("cpu")
 
     text_streamer = TextStreamer(tokenizer)
     output = model.generate(**inputs, streamer=text_streamer, max_new_tokens=max_new_tokens)