I want to know what is the meaning of some items in the log output by distributed training

[Socrates1029]

I use the script "distributed_train.sh" to train the model and get some output logs.

Train: 4 [ 96/97 (100%)] Loss: 2.157 (2.13) Time: 1.024s, 500.22/s (1.020s, 501.91/s) LR: 8.333e-04 Data: 0.000 (0.006)

I want to know what the "Time" and "Data" in the above line mean.

Test: [ 19/19] Time: 0.197 (0.351) Loss: 0.9819 (1.5991) Acc@1: 79.4118 (45.0400) Acc@5: 97.0588 (89.4900)

I want to know the meaning of the brackets in the above two lines of logs, for example, Time: 1.024s, 500.22/s (1.020s, 501.91/s), Loss: 0.9819 (1.5991) and Acc@1: 79.4118 (45.0400), etc. .

Your explanation is greatly appreciated!