update description

app.py

@@ -105,7 +105,7 @@ title = "# Vietnamese Streaming RNN-T"
 description = """
 Visit <https://github.com/HKAB/vietnamese-rnnt-tutorial/> for more information.
 
-- This model runs on a very slow CPU (it's Free tier) so the RTF of FP32 model is around 1.5.
+- This model runs on a very slow CPU (it's Free tier) so the RTF of FP32 model is around 1.5 (which means it will take 1.5 times the duration of the audio to process it).
 - This model mights not work with your microphone since it was trained on a quite clean dataset. Try to speak loudly 😃
 - Although you upload a full audio file, the model will process it in a streaming fashion.
 """