How to train ResNet50 for ImageNet1k

I was taking the reference from the example for th3 ResNet50v1.5, it contains some script for the training and quick guide. I run the script `python ./multiproc.py --nproc_per_node 8 ./main.py --arch resnet50 -c fanin --label-smoothing 0.1 <path to imagenet>`. Is it the true script to train the model from scratch?

Also running this raises an error for image decoder