Update app.py

app.py

@@ -8,7 +8,9 @@ device = 'cuda' if torch.cuda.is_available() else "cpu"
 # torch.set_default_device("cuda")
 
 # Initialize the model and tokenizer
-model = AutoModelForCausalLM.from_pretrained("ManishThota/Sparrow").to(device)
+model = AutoModelForCausalLM.from_pretrained("ManishThota/Sparrow",  torch_dtype=torch.float16, 
+    device_map="auto",
+    trust_remote_code=True).to(device)
 tokenizer = AutoTokenizer.from_pretrained("ManishThota/Sparrow", trust_remote_code=True)
 
 def predict_answer(image, question):