Update app.py

app.py

@@ -15,9 +15,11 @@ device: str = "cuda" if torch.cuda.is_available() else "cpu"
 @spaces.GPU
 def inference(audio_file: str, model_name: str, vocals: bool, drums: bool, bass: bool, other: bool, mp3: bool, mp3_bitrate: int) -> Tuple[str, gr.HTML]:
     separator = get_model(name=model_name)
+    log_messages = []
 
     def stream_log(message):
-        return f"<pre style='margin-bottom: 0;'>[{model_name}] {message}</pre>"
+        log_messages.append(f"[{model_name}] {message}")
+        return gr.HTML("<pre style='margin-bottom: 0;'>" + "<br>".join(log_messages) + "</pre>")
 
     yield None, stream_log("Starting separation process...")
     yield None, stream_log(f"Loading audio file: {audio_file}")
@@ -47,13 +49,18 @@ def inference(audio_file: str, model_name: str, vocals: bool, drums: bool, bass:
     ref = wav.mean(0)
     wav = (wav - ref.view(1, -1))
     yield None, stream_log("Audio loaded successfully. Applying model...")
-    
+
     # Use apply_model as a standalone function
     try:
         sources = apply_model(separator, wav.to(device), device=device)
-    except Exception as e:
+    except ValueError as e:
         yield None, stream_log(f"Error applying model: {str(e)}")
-        raise gr.Error(f"Failed to apply model: {str(e)}")
+        yield None, stream_log(f"Separator sources: {separator.sources}")
+        yield None, stream_log(f"WAV shape: {wav.shape}")
+        raise gr.Error(f"Failed to apply model: {str(e)}. This might be due to incompatible audio format or model configuration.")
+    except Exception as e:
+        yield None, stream_log(f"Unexpected error applying model: {str(e)}")
+        raise gr.Error(f"An unexpected error occurred while applying the model: {str(e)}")
 
     # Process the sources
     sources = [source * ref.view(1, -1) + ref.view(1, -1) for source in sources]