I dont understand the Mask-RCNN training output

I just switched from TensorFlow to PyTorch, so I know my neural networks, but not so much PyTorch.

When I do a training with Mask R-CNN (using ResNet-50) I get this kind of output:

Epoch: [8] [410/454] eta: 0:00:29 lr: 0.005000 loss: 0.4226 (0.4910) loss_classifier: 0.1969 (0.2415) loss_box_reg: 0.1210 (0.1449) loss_mask: 0.0751 (0.0860) loss_objectness: 0.0022 (0.0045) loss_rpn_box_reg: 0.0064 (0.0141) time: 0.6515 data: 0.4290 max mem: 1635

I know these are different losses, so they should go down, ok.
But what kind of losses?
And why are there 2 numbers for each loss?

And what is meant with “data”?

Can someone explain these values to me?
I tried to find some documentation about this but failed (maybe someone should document that!?).

Someone flagged this question so it got hidden.
Please tell me what is wrong with it and I revise my question.