Update app.py

app.py

@@ -26,7 +26,7 @@ cancel_encode = False
 cancel_decode = False
 cancel_stream = False
 
-@spaces.GPU(duration=250)  # Changed GPU duration to 250 seconds
+@spaces.GPU(duration=250)
 def encode_audio(audio_file_path):
     global cancel_encode
     try:
@@ -38,6 +38,9 @@ def encode_audio(audio_file_path):
         with torch.no_grad():
             tokens = semanticodec.encode(audio)
 
+        # Debugging print statement
+        print(f"Tokens shape: {tokens.shape}, dtype: {tokens.dtype}")
+
         # Convert to NumPy and save to a temporary .owie file
         tokens_numpy = tokens.detach().cpu().numpy()
         temp_fd, temp_file_path = tempfile.mkstemp(suffix=".owie")
@@ -52,12 +55,13 @@ def encode_audio(audio_file_path):
         return temp_file_path
 
     except Exception as e:
+        print(f"Encoding error: {e}")
         return f"Encoding error: {e}"
 
     finally:
         cancel_encode = False  # Reset cancel flag after encoding
 
-@spaces.GPU(duration=250)  # Changed GPU duration to 250 seconds
+@spaces.GPU(duration=250)
 def decode_audio(encoded_file_path):
     global cancel_decode
     try:
@@ -79,12 +83,13 @@ def decode_audio(encoded_file_path):
         return temp_wav_path
 
     except Exception as e:
+        print(f"Decoding error: {e}")
         return f"Decoding error: {e}"
 
     finally:
         cancel_decode = False  # Reset cancel flag after decoding
 
-@spaces.GPU(duration=250)  # Changed GPU duration to 250 seconds
+@spaces.GPU(duration=250)
 async def stream_decode_audio(encoded_file_path) -> Generator[tuple, None, None]:
     global cancel_stream
     try:
@@ -122,27 +127,27 @@ with gr.Blocks() as demo:
     gr.Markdown("## Audio Compression with SemantiCodec (GPU/CPU)")
 
     with gr.Tab("Encode"):
-        input_audio = gr.Audio(label="Input Audio", type="filepath")
+        input_audio = gr.Audio(label="Input Audio", type="filepath")  # Using "filepath" mode
         encode_button = gr.Button("Encode")
         cancel_encode_button = gr.Button("Cancel")
-        encoded_output = gr.File(label="Encoded File (.owie)", type="filepath")
+        encoded_output = gr.File(label="Encoded File (.owie)", type="filepath")  # Using "filepath" mode
 
         encode_button.click(encode_audio, inputs=input_audio, outputs=encoded_output)
         cancel_encode_button.click(lambda: globals().update(cancel_encode=True),
                                      outputs=None)  # Set cancel_encode flag
 
     with gr.Tab("Decode"):
-        input_encoded = gr.File(label="Encoded File (.owie)", type="filepath")
+        input_encoded = gr.File(label="Encoded File (.owie)", type="filepath")  # Using "filepath" mode
         decode_button = gr.Button("Decode")
         cancel_decode_button = gr.Button("Cancel")
-        decoded_output = gr.Audio(label="Decoded Audio", type="filepath")
+        decoded_output = gr.Audio(label="Decoded Audio", type="filepath")  # Using "filepath" mode
 
         decode_button.click(decode_audio, inputs=input_encoded, outputs=decoded_output)
         cancel_decode_button.click(lambda: globals().update(cancel_decode=True),
                                      outputs=None)  # Set cancel_decode flag
 
     with gr.Tab("Streaming"):
-        input_encoded_stream = gr.File(label="Encoded File (.owie)", type="filepath")
+        input_encoded_stream = gr.File(label="Encoded File (.owie)", type="filepath")  # Using "filepath" mode
         stream_button = gr.Button("Start Streaming")
         cancel_stream_button = gr.Button("Cancel")
         audio_output = gr.Audio(label="Streaming Audio Output", streaming=True)