Update app.py

app.py

@@ -2,6 +2,7 @@ import torch
 
 import numpy as np
 import gradio as gr
+import soundfile as sf
 
 from transformers import pipeline
 from huggingface_hub import InferenceClient
@@ -30,9 +31,11 @@ def generate_audio(text,):
     prompt = f"Take the next sentence and enrich it with details. Keep it compact. {text}"
     output = client.text_generation(prompt, max_new_tokens=100)
     out = vibes(output)
-    np_audio = out["audio"][0]
+
+    with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as f:
+        sf.write(f.name, out["audio"][0].T, out["sampling_rate"])
     
-    return np_audio
+    return f.name
 
 css = """
 #container{
@@ -56,7 +59,7 @@ with gr.Blocks(css=css) as demo_blocks:
             btn = gr.Button("Generate Music!🎶")
             
         with gr.Column():
-            out = gr.Audio(type="numpy", autoplay=False, label=f"Generated Music", show_label=True,)
+            out = gr.Audio(autoplay=False, label=f"Generated Music", show_label=True,)
 
 
     with gr.Accordion("Run MusicGen with Transformers 🤗", open=False):