U

app.py

@@ -35,10 +35,18 @@ def model_inference(input_wav, language):
 
 
 def launch():
-
     with gr.Blocks() as demo:
         gr.Markdown("# Cantonese Call Transcriber")
-        gr.Markdown("Click the example at the bottom to try it out.")
+        gr.Markdown(
+            """
+        This tool transcribes Cantonese audio calls into text.
+        
+        ## How to use:
+        1. Upload an audio file or use the provided example.
+        2. Click the 'Process Audio' button.
+        3. The transcription will appear in the output box.
+        """
+        )
 
         # Define components
         audio_input = gr.Audio(label="Input")