Update app.py

app.py

@@ -35,9 +35,10 @@ print(f"Low memory: {LOW_MEMORY}")
 
 # Quantization configuration for efficient model loading
 # Define BitsAndBytesConfig
-quantization_config = BitsAndBytesConfig(load_in_4bit=True,
-                                bnb_4bit_quant_type="nf4",
-                                bnb_4bit_compute_dtype=torch.float16)
+quantization_config = BitsAndBytesConfig(
+    load_in_4bit=True,
+    bnb_4bit_compute_dtype=torch.float16
+)
 
 
 # Load models only once