OSError: cannot identify image file <_io.BufferedReader name='./data/LSUN_fix/LSUN_fix/._correct_resize_1.png'>

[zwy1996]

Hi, nice work!

When I ran experiments on cifar10, I got the error, it seemed like that the files of LSUN_fix. I was a freshman on the task, Would you please give me some advice?

Best wished to you!

[ksaito-ut]

Hi, thanks for your interest in our work.
It is probably because the directory contains hidden files, ._correct_resize_1.png, .correct_resize_2.png, ....
Can you just delete the images starting from . in the LSUN directory?
If my memory is correct, I tried that way.

[zwy1996]

Thank you very much!

Now, I ran the experiments normally!

Best wishes to you!

[zwy1996]

I was very sorry to bother you.

Now I got another error. I was a little confused abount it. Because, it happened suddenly

Traceback (most recent call last):
File "main.py", line 105, in
main()
File "main.py", line 96, in main
ood_loaders, model, optimizer, ema_model, scheduler)
File "/data1/aux_task/openset/OP_Match-main/trainer.py", line 96, in train
sampler = train_sampler(unlabeled_dataset),
File "/home/anaconda3/envs/openset/lib/python3.6/site-packages/torch/utils/data/sampler.py", line 94, in init
"value, but got num_samples={}".format(self.num_samples))
ValueError: num_samples should be a positive integer value, but got num_samples=0
0%| | 0/1024 [00:20<?, ?it/s]

[ksaito-ut]

It is because unlabeled_dataset has no samples due to exclude_dataset(args, unlabeled_dataset, ema_model.ema). Did you change some hyper-parameters to run the experiments? If you change it, an error can happen. If you do not change it, I need to look into the code to figure out the cause. I did not encounter the error in my environment. Best, Kuniaki 2021年10月23日(土) 4:10 zwy1996 ***@***.***>:

…

I was very sorry to bother you. Now I got another error. I was a little confused abount it. Because, it happened suddenly Traceback (most recent call last): File "main.py", line 105, in main() File "main.py", line 96, in main ood_loaders, model, optimizer, ema_model, scheduler) File "/data1/aux_task/openset/OP_Match-main/trainer.py", line 96, in train sampler = train_sampler(unlabeled_dataset), File "/home/anaconda3/envs/openset/lib/python3.6/site-packages/torch/utils/data/sampler.py", line 94, in *init* "value, but got num_samples={}".format(self.num_samples)) ValueError: num_samples should be a positive integer value, but got num_samples=0 0%| | 0/1024 [00:20<?, ?it/s] — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#2 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AFOP4APH6YTM4RIQ47ABZJ3UIJUW7ANCNFSM5GOJFUVQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>.

-- Kuniaki Saito

[zwy1996]

Hi,

I did not change parameter, just used the default parameter.

[ksaito-ut]

And, which dataset you used? cifar10? Kuniaki 2021年10月24日(日) 0:59 zwy1996 ***@***.***>:

…

Hi, I did not change parameter, just used the default parameter. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#2 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AFOP4AKWT27Z2645YJQ4DTLUIOHDPANCNFSM5GOJFUVQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>.

-- Kuniaki Saito

[zwy1996]

yes， cifar 10
The instruction was "sh run_cifar10.sh 50 output 1"

[ksaito-ut]

Can you put a whole log? Also, I asked my colleague to run the code, he ran it without using amp, I confirmed that it worked well. I do not think amp hurts performance since I used it for all experiments, but disabling amp may be worth trying. Kuniaki 2021年10月26日(火) 10:09 zwy1996 ***@***.***>:

…

yes， cifar 10 The instruction was "sh run_cifar10.sh 50 output 1" — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#2 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AFOP4AOAN34D6Q6HUYFUPB3UI2ZCBANCNFSM5GOJFUVQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>.

-- Kuniaki Saito

[zwy1996]

ok, I will try again!

Thank you very much!

[zwy1996]

Hi,

It was too strange，I ran the experiments normally now, I had not met the error anymore, maybe I had some incorrect manner.

I was very sorry to bother you! Thank you very much!