BentoML

Hi Bharath, if the GPU is available then it will be used by the runner automatically

<@U023HR5862J> its throwing runtime error as below
```RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cpu and cuda:0! (when 
checking argument for argument index in method wrapper_CUDA__index_select)```

_input is
```{'input_ids': tensor([[   259,  74606, 207850,  74295]], device='cuda:0'), 'attention_mask': tensor([[1, 1, 1, 1]], device='cuda:0')}```
which is using cuda but model_runner is not using cuda.

i Bharath, could you post your service.py? If you are using `bentoml.transformers` it should automatically lift tensor to cuda. Your error message means some tensor are lifted to cuda but some are not. Very strange behavior.