SRFlow icon indicating copy to clipboard operation
SRFlow copied to clipboard

Published 20 hours ago • verified publisher icon andreas128

train error: svd_cuda

Open vkl-git opened this issue 4 years ago • 4 comments
trafficstars

Hello, thank you for your work.When I tried to train with my own datasets, I used the SRFlow_DF2K_4x model, but during the training process, "svd_cuda: the updating process of SBDSDC did not converage (error:11)" error appeared.Have you ever encountered the same reason in your training?Is there a solution? image

vkl-git avatar Feb 20 '21 07:02 vkl-git

which cuda and pytorch version?

flybiubiu avatar Feb 20 '21 14:02 flybiubiu

Thank you for your response.  I used CUDA 10. 0 and pytorch1.7. Are there any problems? 

---Original--- From: "flybiubiu"<[email protected]> Date: Sat, Feb 20, 2021 22:20 PM To: "andreas128/SRFlow"<[email protected]>; Cc: "viki-li"<[email protected]>;"Author"<[email protected]>; Subject: Re: [andreas128/SRFlow] train error: svd_cuda (#21)

which cuda and pytorch version?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.

vkl-git avatar Feb 20 '21 14:02 vkl-git

I train it with same problem before, But I run setup file all problem are display. If GPU is 3090,cuda version should be cuda 11.0

flybiubiu avatar Feb 20 '21 14:02 flybiubiu

Hi viki-li, this error usually occurs in the back propagation step of the 1x1 convolution, when the learning rate is too high.

Does it help to reduce the learning rate?

andreas128 avatar Feb 22 '21 11:02 andreas128