Update app.py

app.py

@@ -1,21 +1,61 @@
+!pip install -q -U diffusers transformers accelerate
+     
+
+from diffusers import AudioLDM2Pipeline
+     
+
+import torch
+model_id = "cvssp/audioldm2"
+
+pipe = AudioLDM2Pipeline.from_pretrained(
+    model_id,
+    torch_dtype = torch.float16,
+).to("cuda")
+     
+
+
+prompt = "A cheerful ukulele strumming in a beachside jam."
+     
+
+generator = torch.Generator("cuda").manual_seed(0)
+     
+
+audio = pipe(
+    prompt,
+    audio_length_in_s = 10.24,
+    generator = generator,
+).audios[0]
+     
+
+audio = pipe(
+    prompt,
+    negative_prompt = negative_prompt,
+    audio_length_in_s = 10.24,
+    generator = generator,
+).audios[0]
+     
+from IPython.display import Audio
+Audio(audio, rate = 16000)
+     
+!pip install -q gradio
 import gradio as gr
-from transformers import pipeline
-
-# Load the text-to-audio model
-tts = pipeline("create_music", model="facebook/musicgen-small")
-
-def text_to_audio(text):
-    output = tts(text)
-    return output["audio"]
-
-# Create Gradio interface
-iface = gr.Interface(
-    fn=create_music,
-    inputs=gr.Textbox(label="Enter Text"),
-    outputs="audio"
-    title="Text-to-Audio Chatbot",
-    description="Enter text, and get an AI-generated audio response."
-)
-
-if __name__ == "__main__":
-    iface.launch()
+     
+
+def create_music(prompt):
+  negative_prompt = "Low quality"
+  audio = pipe(
+    prompt,
+    negative_prompt = negative_prompt,
+    audio_length_in_s = 10.24,
+    generator = generator,
+  ).audios[0]
+  return 16000, audio
+     
+
+interface = gr.Interface(
+    title = "Music Generation App",
+    examples = ["A cheerful ukulele strumming in a beachside jam."],
+    fn = create_music,
+    inputs = gr.Textbox(),
+    outputs = "audio",
+).launch(debug=True, share=True)