Get values return by Tuple Object in Maskcrnn libtorch

I’m a newbie in C++ and libtorch, I try load model by torchscript and execute inference and result like this:

Load model successful!
[W] Warning: RCNN always returns a (Losses, Detections) tuple in scripting (function )
Inference time: 2321 ms
({}, [{boxes: [ CPUFloatType{0,4} ], labels: [ CPULongType{0} ], scores: [ CPUFloatType{0} ], masks: [ CPUFloatType{0,1,800,800} ]}])

How do I get value boxes, labels, scores and masks from return output object using c++ ?
I tried many ways but compile always error with “c10::IValue” error thrown
One more question, why is the time inference when I convert the model to torchscript, executed by C++ is slower than python ?
Many thanks