search

Li-Chongyi / Dehamer

MIT License
70 stars 11 forks source link

Some questions about experimental settings. #2

Open Hi-WenXin opened 2 years ago

Hi-WenXin commented 2 years ago

In paper 4.1 Experimental Settings, I found that when training the network the image patch is from small to full size, but I do not see the corresponding part in the code. Will this affect the training of the model? And the batch size is 2, which maybe is too tiny for V100 GPU. Why did you set such a small value?

Li-Chongyi commented 2 years ago

Hi, You can manually set the patch size in the different training stages, ranging from small to large. We found this kind of training strick can improve the performance.

It is ok if you use other sizes of the batch.

Hi-WenXin @.***> 于2022年7月16日周六 14:43写道：

In paper 4.1 Experimental Settings, I found that when training the network the image patch is from small to full size, but I do not see the corresponding part in the code. Will this affect the training of the model? And the batch size is 2, which maybe is too tiny for V100 GPU. Why did you set such a small value?

— Reply to this email directly, view it on GitHub https://github.com/Li-Chongyi/Dehamer/issues/2, or unsubscribe https://github.com/notifications/unsubscribe-auth/AIEVQEISCU25ILYYPU3UXPDVUJKYXANCNFSM53XTJLUQ . You are receiving this because you are subscribed to this thread.Message ID: @.***>

Hi-WenXin commented 2 years ago

Thanks so much.

TaLe97TJU commented 1 year ago

您好，读了您的论文受益匪浅，但是最近复现Dehamer遇到了一点问题，请问可否提供一下您手动调整patch size的具体数据，比如在第几个epoch，patch size 调整为多少？