Update app.py

app.py

@@ -47,18 +47,13 @@ ocr_model = ocr_predictor(
 
 
 if torch.cuda.is_available():
-    # Load the processor (no need to move to GPU)
     processor = LlavaNextProcessor.from_pretrained("llava-hf/llava-v1.6-mistral-7b-hf")
-
-    # Load the model and move it to GPU
     vision_model = LlavaNextForConditionalGeneration.from_pretrained(
         "llava-hf/llava-v1.6-mistral-7b-hf",
         torch_dtype=torch.float16,
         low_cpu_mem_usage=True,
         load_in_4bit=True,
-    ).to("cuda")  # Move the model to GPU
-
-
+    )
 
 
 @spaces.GPU