Update app.py

app.py

@@ -67,6 +67,8 @@ def decode(tokenizer, encoding):
 def update_options(input, num_tokens):
     tokenized_prompts = tokenizer.encode([input], True, False)
     print("Processed prompt")
+    model.model.to("cuda")
+    model.model.device = "cuda"
     alive_gens, _ = model.sup_generate(prompt_tokens=tokenized_prompts, 
                                             smoothing="geom",
                                             max_gen_len=num_tokens,