Regarding the issue of continuous memory growth during the training process

[Daming-wang]

When I run the ’/mlcommons/training/image_classification/tensorflow/resnet_ctl_imagenet_main.py' script, my computer's memory keeps increasing, and by the 7th epoch, the memory usage has already reached 35GB. Is this phenomenon normal?

[ShriyaPalsamudram]

Sorry but the resnet50 benchmark is dropped from the training benchmarks suite so this issue cannot be addressed at this time.