Add cuda instruction

README.md

@@ -121,7 +121,8 @@ You can deploy agents on your varying compute resources as long as it has access
 
   To deploy the ClearML with a GPU allocated add the --gpus argument
    - Install CUDA https://docs.nvidia.com/cuda/cuda-installation-guide-linux/index.html
-   - install pytorch with cuda support https://pytorch.org/get-started/locally with Cuda support 
+   - install pytorch with cuda support https://pytorch.org/get-started/locally with Cuda support
+   - use cuda as device in script
    
    - clearml-agent daemon --queue default --gpus 0