Update app.py

app.py

@@ -56,7 +56,7 @@ def stream(audio, chunk_length_s):
         chunk_end_time = end_pos / sampling_rate
 
         # Save timestamps for the current chunk
-        timestamps.append((chunk_start_time, chunk_end_time))
+        timestamps.append((round(chunk_start_time, 2), round(chunk_end_time, 2)))
 
         chunk = array[start_pos:end_pos]
         chunk_mp3 = numpy_to_mp3(chunk, sampling_rate=sampling_rate)
@@ -65,7 +65,7 @@ def stream(audio, chunk_length_s):
             first_time = round(time.time() - start_time, 2)
         run_time = round(time.time() - start_time, 2)
 
-        yield chunk_mp3, first_time, run_time
+        yield chunk_mp3, first_time, run_time, timestamps
 
     # Save timestamps to a CSV file
     with open("timestamps_30sec.csv", mode="w", newline="") as file:
@@ -82,14 +82,26 @@ def stream(audio, chunk_length_s):
 with gr.Blocks() as demo:
     with gr.Row():
         with gr.Column():
-            audio_in = gr.Audio(value="librispeech.wav", sources=["upload"], type="numpy")
+            audio_in = gr.Audio(value="librispeech.wav", sources=["upload"], type="numpy", label="Input Audio")
             chunk_length = gr.Slider(minimum=2, maximum=10, value=2, step=2, label="Chunk length (s)")
             run_button = gr.Button("Stream audio")
         with gr.Column():
-            audio_out = gr.Audio(streaming=True, autoplay=True, format="mp3", label="mp3")
+            audio_out = gr.Audio(streaming=True, autoplay=True, format="mp3", label="Streamed MP3 Audio")
             first_time = gr.Textbox(label="Time to first chunk (s)")
             run_time = gr.Textbox(label="Time to current chunk (s)")
 
-    run_button.click(fn=stream, inputs=[audio_in, chunk_length], outputs=[audio_out, first_time, run_time])
+    with gr.Row():
+        timestamps_output = gr.Dataframe(
+            headers=["Start Time (s)", "End Time (s)"],
+            label="Timestamps for the First 30 Seconds",
+            interactive=False
+        )
+
+    # Updated function outputs with the timestamps table
+    run_button.click(
+        fn=stream,
+        inputs=[audio_in, chunk_length],
+        outputs=[audio_out, first_time, run_time, timestamps_output]
+    )
 
 demo.launch()