[Enhancement] The memory filed is not saved in all visualization backends.

[HAOCHENYE]

The memory filed is not saved in all visualization backends. Take LocalVisbackend as an example, the saved json turns out like this:

{"lr": 0.001, "data_time": 0.008973455429077149, "loss": 5.027596592903137, "time": 0.2356168508529663, "epoch": 1, "step": 10}
{"lr": 0.001, "data_time": 0.007732248306274414, "loss": 2.66796715259552, "time": 0.032484245300292966, "epoch": 1, "step": 20}
{"lr": 0.001, "data_time": 0.007663464546203614, "loss": 2.699092149734497, "time": 0.03233048915863037, "epoch": 1, "step": 30}
{"lr": 0.001, "data_time": 0.007588768005371093, "loss": 2.6876137018203736, "time": 0.032196879386901855, "epoch": 1, "step": 40}
{"lr": 0.001, "data_time": 0.007547187805175781, "loss": 2.4950002908706663, "time": 0.032173895835876466, "epoch": 1, "step": 50}
{"lr": 0.001, "data_time": 0.0075434684753417965, "loss": 2.529031848907471, "time": 0.03212928771972656, "epoch": 1, "step": 60}
{"lr": 0.001, "data_time": 0.007540845870971679, "loss": 2.5454934000968934, "time": 0.03212835788726807, "epoch": 1, "step": 70}
{"lr": 0.001, "data_time": 0.007552504539489746, "loss": 2.627494549751282, "time": 0.03256125450134277, "epoch": 1, "step": 80}
{"lr": 0.001, "data_time": 0.007521510124206543, "loss": 2.4862494230270387, "time": 0.03204777240753174, "epoch": 1, "step": 90}
{"lr": 0.001, "data_time": 0.008119916915893555, "loss": 2.6292912483215334, "time": 0.0344351053237915, "epoch": 1, "step": 100}
{"lr": 0.001, "data_time": 0.007765579223632813, "loss": 2.5920512676239014, "time": 0.03332774639129639, "epoch": 1, "step": 110}
{"lr": 0.001, "data_time": 0.0076437234878540036, "loss": 2.684328055381775, "time": 0.03258049488067627, "epoch": 1, "step": 120}
{"lr": 0.001, "data_time": 0.007628417015075684, "loss": 2.9349906206130982, "time": 0.03266925811767578, "epoch": 1, "step": 130}
{"lr": 0.001, "data_time": 0.007643294334411621, "loss": 2.536902570724487, "time": 0.032727813720703124, "epoch": 1, "step": 140}
{"lr": 0.001, "data_time": 0.007609415054321289, "loss": 2.662415337562561, "time": 0.03258848190307617, "epoch": 1, "step": 150}
{"lr": 0.001, "data_time": 0.007627367973327637, "loss": 2.510297155380249, "time": 0.03274567127227783, "epoch": 1, "step": 160}
{"lr": 0.001, "data_time": 0.007639122009277344, "loss": 2.4019510746002197, "time": 0.032790541648864746, "epoch": 1, "step": 170}

it does not store the memory information saved in the log file.

2023/01/10 22:49:57 - mmengine - WARNING - `save_param_scheduler` is True but `self.param_schedulers` is None, so skip saving parameter schedulers
2023/01/10 22:49:57 - mmengine - INFO - Epoch(val) [1][ 10/313]    eta: 0:00:03  time: 0.0119  data_time: 0.0049  memory: 481  
2023/01/10 22:49:57 - mmengine - INFO - Epoch(val) [1][ 20/313]    eta: 0:00:03  time: 0.0115  data_time: 0.0048  memory: 305  
2023/01/10 22:49:57 - mmengine - INFO - Epoch(val) [1][ 30/313]    eta: 0:00:03  time: 0.0118  data_time: 0.0048  memory: 305  
2023/01/10 22:49:57 - mmengine - INFO - Epoch(val) [1][ 40/313]    eta: 0:00:03  time: 0.0115  data_time: 0.0048  memory: 305  
2023/01/10 22:49:58 - mmengine - INFO - Epoch(val) [1][ 50/313]    eta: 0:00:03  time: 0.0116  data_time: 0.0048  memory: 305  
2023/01/10 22:49:58 - mmengine - INFO - Epoch(val) [1][ 60/313]    eta: 0:00:02  time: 0.0115  data_time: 0.0048  memory: 305  
2023/01/10 22:49:58 - mmengine - INFO - Epoch(val) [1][ 70/313]    eta: 0:00:02  time: 0.0115  data_time: 0.0048  memory: 305  
2023/01/10 22:49:58 - mmengine - INFO - Epoch(val) [1][ 80/313]    eta: 0:00:02  time: 0.0115  data_time: 0.0048  memory: 305  
2023/01/10 22:49:58 - mmengine - INFO - Epoch(val) [1][ 90/313]    eta: 0:00:02  time: 0.0115  data_time: 0.0048  memory: 305  
2023/01/10 22:49:58 - mmengine - INFO - Epoch(val) [1][100/313]    eta: 0:00:02  time: 0.0119  data_time: 0.0048  memory: 305  
2023/01/10 22:49:58 - mmengine - INFO - Epoch(val) [1][110/313]    eta: 0:00:02  time: 0.0115  data_time: 0.0048  memory: 305