Update app.py

app.py

@@ -1,23 +1,63 @@
 import gradio as gr
-from huggingface_hub import InferenceClient
-
-# Logic for your app
-def master_audio(input_file):
-    # This is where the magic happens. 
-    # You would typically call a model like SonicMaster here.
-    # For now, we'll return the file to show the UI structure.
-    return input_file
-
-# Create the UI
-with gr.Blocks() as demo:
-    gr.Markdown("# 🎵 AI Pro Studio")
-    gr.Markdown("Upload your demo and let AI polish the EQ and Dynamics.")
+from gradio_client import Client, handle_file
+from transformers import pipeline
+
+# 1. Initialize the Genre Classifier (Local/Lightweight)
+# This model identifies genres like Pop, Rock, Hip-Hop, etc.
+classifier = pipeline("audio-classification", model="MIT/ast-finetuned-audioset")
+
+# 2. Connect to the Mastering Engine
+mastering_client = Client("amaai-lab/SonicMaster")
+
+def auto_master(audio_path):
+    if audio_path is None:
+        return None, "No audio detected."
+
+    # Step 1: Detect Genre
+    # We take a 30-second snippet to speed up detection
+    results = classifier(audio_path)
+    top_genre = results[0]['label']
     
-    with gr.Row():
-        audio_input = gr.Audio(type="filepath", label="Upload Raw Song")
-        audio_output = gr.Audio(label="Mastered Version")
+    # Step 2: Map Genre to a Professional Prompt
+    genre_prompts = {
+        "Hip Hop": "Master this for heavy bass and crisp vocals. High energy.",
+        "Rock": "Give it a warm analog feel with punchy drums and clear guitars.",
+        "Pop": "Make it sound like a radio hit. Bright, loud, and very wide.",
+        "Classical": "Preserve the dynamic range but add professional clarity.",
+        "Electronic": "Focus on sub-bass clarity and sharp high-end transients."
+    }
     
-    btn = gr.Button("Master My Track")
-    btn.click(fn=master_audio, inputs=audio_input, outputs=audio_output)
+    # Fallback if genre isn't in our list
+    final_prompt = genre_prompts.get(top_genre, "Master this song to sound professionally produced.")
+
+    # Step 3: Send to SonicMaster
+    result = mastering_client.predict(
+        audio=handle_file(audio_path),
+        prompt=final_prompt,
+        seed=42,
+        steps=20,
+        cfg_scale=3.5,
+        api_name="/predict"
+    )
+    
+    return result[1], f"Detected Genre: {top_genre} | Applied Style: {final_prompt}"
+
+# 3. Create the "Smart" UI
+with gr.Blocks(theme=gr.themes.Monochrome()) as demo:
+    gr.Markdown("# 🤖 Smart AI Mastering Studio")
+    gr.Markdown("Upload a track. We'll identify the genre and apply custom studio settings.")
+    
+    with gr.Row():
+        input_audio = gr.Audio(label="Raw Audio File", type="filepath")
+        output_audio = gr.Audio(label="AI Mastered Output")
+        
+    status_msg = gr.Textbox(label="Analysis Status", interactive=False)
+    run_btn = gr.Button("✨ Auto-Master My Track", variant="primary")
+
+    run_btn.click(
+        fn=auto_master, 
+        inputs=input_audio, 
+        outputs=[output_audio, status_msg]
+    )
 
 demo.launch()