Update src/app/model.py

src/app/model.py

@@ -40,14 +40,13 @@ def load_model_tokenizer_and_processor(model_name: str, device: str) -> Any:
             torch_dtype=torch.bfloat16,
             token=access_token
         )
-        model = model.to(device=device)
+        model = model.eval().to(device=device)
         tokenizer = AutoTokenizer.from_pretrained(
             model_name, trust_remote_code=True, token=access_token
         )
         processor = AutoProcessor.from_pretrained(
             model_name, trust_remote_code=True, token=access_token
         )
-        model.eval()
 
         # Log the successful loading of the model, tokenizer and processor
         logging.info("Model, tokenizer and processor loaded successfully.")