updates for qwen specific items

app.py

@@ -9,7 +9,7 @@ import requests
 from io import BytesIO
 from urllib.request import urlopen, Request
 import gc
-from transformers import AutoProcessor, AutoModelForCausalLM
+from transformers import AutoProcessor, Qwen2AudioForConditionalGeneration
 
 # Configure logging
 logging.basicConfig(
@@ -51,9 +51,9 @@ def load_model():
         processor = AutoProcessor.from_pretrained(MODEL_ID)
         logger.info("Processor loaded successfully")
         
-        # Load model with basic FP16 config
+        # Load model with basic FP16 config - using the correct model class
         logger.info(f"Loading model from {MODEL_ID}")
-        model = AutoModelForCausalLM.from_pretrained(
+        model = Qwen2AudioForConditionalGeneration.from_pretrained(
             MODEL_ID,
             torch_dtype=torch.float16,
             device_map="auto",