Update app.py

app.py

@@ -13,7 +13,7 @@ device = "cuda" if torch.cuda.is_available() else "cpu"
 print(f"Using device: {device}")
 
 # Load the Byaldi and Qwen2-VL models
-rag_model = RAGMultiModalModel.from_pretrained("vidore/colpali").to(device)  # Move Byaldi to GPU
+rag_model = RAGMultiModalModel.from_pretrained("vidore/colpali")  # Do not move Byaldi to GPU
 qwen_model = Qwen2VLForConditionalGeneration.from_pretrained(
     "Qwen/Qwen2-VL-7B-Instruct", trust_remote_code=True, torch_dtype=torch.bfloat16
 ).to(device)  # Move Qwen2-VL to GPU