write

app.py

@@ -52,4 +52,5 @@ if question:
     model_inputs = inputs.to('cuda')
     generated_ids = model.generate(**model_inputs, max_new_tokens=300, do_sample=True, pad_token_id=tokenizer.eos_token_id)
     decoded_output = tokenizer.batch_decode(generated_ids)
+    st.write(decoded_output)