Update app.py

app.py

@@ -11,7 +11,7 @@ from vllm import LLM, SamplingParams
 def load_model_processor(model_path):
     processor = AutoProcessor.from_pretrained(model_path)
     llm = LLM(
-        model=model_path, trust_remote_code=True, gpu_memory_utilization=0.4,  
+        model=model_path, trust_remote_code=True, gpu_memory_utilization=0.8,  
         enforce_eager=True,  device = "cuda",
         limit_mm_per_prompt={"audio": 5},
     )