Handle errors from NVL more gracefully.

[bzamecnik]

Logs:

  File "/usr/local/lib/python2.7/dist-packages/nvgpu/list_gpus.py", line 60, in device_statuses
    return [device_status(device_index) for device_index in range(device_count)]
  File "/usr/local/lib/python2.7/dist-packages/nvgpu/list_gpus.py", line 14, in device_status
    handle = nv.nvmlDeviceGetHandleByIndex(device_index)
  File "/usr/local/lib/python2.7/dist-packages/pynvml.py", line 946, in nvmlDeviceGetHandleByIndex
    _nvmlCheckReturn(ret)
  File "/usr/local/lib/python2.7/dist-packages/pynvml.py", line 405, in _nvmlCheckReturn
    raise NVMLError(ret)
NVMLError_GpuIsLost: GPU is lost

It fails with 500, but should better report the error NVMLError_GpuIsLost: GPU is lost.