coco-hand and coco basic training command difference

[YoungjaeDev]

https://github.com/uyzhang/yolov5_prune#steps

In basic training number 2, coco-hand is based on pre-trained weight, but it is not coco, which one is correct?
As far as I know, it is related to the botleneck ignore problem for sparsity training, should we proceed with the existing pretrained weight before sparsity training?

[uyzhang]

It's been modified. The basic training of data sets does not need to load pre-training weights. thx!

[YoungjaeDev]

@uyzhang is there any reason why you did it with adamw? There is not much difference from sgd to calc sparsity, but if there is another reason, can I know?

[uyzhang]

I tried sparse training with SGD, but the effect was not good. AdamW works well, so I end up using AdamW.