Update app.py

app.py

@@ -183,10 +183,9 @@ def load_models_and_documents():
             max_model_len=12288,
             trust_remote_code=True,
             enforce_eager=True,
-            quantization="bitsandbytes",
+            quantization="awq",
             gpu_memory_utilization=0.9,
-            dtype='auto',
-            load_format="bitsandbytes"
+            dtype='auto'
         )
         model = models.VLLM(llm)