added elastic model

app.py

@@ -48,7 +48,7 @@ import os
 # setup_flash_attention()
 
 # Load model and processor
-# @gr.cache()
+@gr.cache()
 def load_model():
     """Load the musicgen model and processor"""
     processor = AutoProcessor.from_pretrained("facebook/musicgen-large")
@@ -71,8 +71,7 @@ def generate_music(text_prompt, duration=10, temperature=1.0, top_k=250, top_p=0
             text=[text_prompt],
             padding=True,
             return_tensors="pt",
-        )
-        
+        ).to("cuda")
         # Generate audio
         with torch.no_grad():
             audio_values = model.generate(