NTIRE 2022 Efficient Super-Resolution Challenge Forum

Go back to competition Back to thread list Post in this thread

> Could you please give me a hand? (About submittion)

I tried to submit my work three times. However, (compressed 100 images which were gotten by my model) + (the same as first try, but added the readme.txt in it) + (the same as second try, but added my model in it), all the tries submitted on the workshop didn't work.
These 100 images are SR from [ DIV2K_valid_LR_bicubic ] , list as [ 0801x4.png, 0802x4.png, ...]
Could you please tell me how to submit my model? I'd like to have a try!

Posted by: Ssakura-go @ Feb. 15, 2022, 5:33 p.m.

Have you figured out how to submit the results successfully? I'm facing the same issue:(

Posted by: RSGao @ Feb. 18, 2022, 5:26 a.m.

I am facing the same issue plus the limit of the zip file seems 30Mb however 100 Super Resolved Image is about 350Mb

Posted by: deepernewbie @ Feb. 23, 2022, 9:42 a.m.

I am facing the same issue plus the limit of the zip file seems 30Mb however 100 Super Resolved Image is about 350Mb

Posted by: xueqingfand @ Feb. 24, 2022, 3:35 a.m.

Did you figure out how to submit yet? I am also facing this problem. Could anyone who submit successfully give any advice?

Posted by: wyboooo @ Feb. 27, 2022, 6:10 a.m.

https://github.com/codalab/codalab-competitions/issues/3110

Didayolo commented 4 days ago
I think the hard limit for submission is currently 300MB.

There is no error message when trying to upload the submission?

Posted by: xueqingfand @ Feb. 28, 2022, 5:35 a.m.

I could be able to do it after a dozen of trials unfortunately the server is not smooth. Keep trying change the internet service provider and try again (maybe try with your phone etc. I did it over my phone with 4G of the cell phone provider not with my regular internet but maybe that is just a coincidence) there is no limit I could be able to upload a 370MB zip file

Posted by: deepernewbie @ Feb. 28, 2022, 10:41 a.m.

Hi, everyone! Sorry to see the same question posed by you all.
The right submission is as following:
① the 100 image after your training model; (NOTE: the name of images should be same as 100 inputs, namely the image in /DIV2K/DIV2K_valid_LR_bicubic/X4/*.png)
② add the readme.txt, for example:
Runtime per image [s] : 0.10
Parameters : 433448
Extra Data [1] / No Extra Data[0] : 1
Other description : Model name : rfdn_4_model ; GPU : Nvidia RTX 2080;Extra data : Flickr2K

over~ Good luck!

Posted by: Ssakura-go @ March 10, 2022, 9:53 p.m.
Post in this thread