fix: description

app.py

@@ -9,7 +9,7 @@ from transformers.pipelines.audio_utils import ffmpeg_read
 MODEL_NAME = "jensenlwt/whisper-small-singlish-122k"
 FILE_LIMIT_MB = 1000
 
-device = 0 if torch.cuda.is_available() else "cpu"
+device = "cuda:0" if torch.cuda.is_available() else "cpu"
 
 pipe = pipeline(
     task="automatic-speech-recognition",
@@ -59,7 +59,9 @@ file_transcribe = gr.Interface(
     layout="horizontal",
     theme="huggingface",
     title="Whisper Small: Singlish Edition 🇸🇬",
-    description=(""),
+    description=(
+        "NOTE: Current space seems to cut off the last few seconds of the recording. For exploration, I would recommend sticking to audio <10s long."
+    ),
     allow_flagging="never",
 )