OutOfRangeError (see above for traceback): RandomShuffleQueue '_2_shuffle_batch/random_shuffle_queue' is closed and has insufficient elements (requested 1, current size 0) #5
Comments
|
Even I got the same error python -m cyclegan.main --to_train=1 --log_dir=cyclegan/output/cyclegan/exp_01 --config_filename=cyclegan/configs/exp_01.json Caused by op u'shuffle_batch', defined at: OutOfRangeError (see above for traceback): RandomShuffleQueue '_1_shuffle_batch/random_shuffle_queue' is closed and has insufficient elements (requested 1, current size 0) |
|
@ambarishgurjar Same to you. Have you deal with it? |
|
It was really frustrating, I tried downgrading the tensorflow version and it still didn't work. Finally I downloaded another implementation of the same project https://github.com/architrathore/CycleGAN/ It is the same algorithm but slightly simplified. You may encounter some problems in this one too If you need the exact changes in code then I will give it to you once I go to my office on Monday. All my data is stored in the work PC |
|
@ambarishgurjar Thanks a lot !! |
ghost
mentioned this issue
|
I'm having the same type of error even without shuffling. Problem is i really need this specific implementation because I'm trying to run an implementation that adds attention capabilities to the GAN and is based off of this one. I was hoping to solve with a downgrade, but from what I gain it doesn't work. How far did you rollback? I was thinking to try tf 0.8. |
|
I solved the issue. "create_cyclegan_dataset.py" creates .csv files that end with this sequence of characters to separate lines : "\r\r\n" and that causes errors during parsing that lead to files not being loaded at all thus making the queues empty. If this sequence of characters is replaced with a simple "\n" character in any text editor, the csv is parsed correctly and everything is back to working order. |
|
@ambarishgurjar Hi,I have met the same issue as you Have you figured it out ?I am really confused,thanks a lot |
|
Hey @heleibin have you tried removing the characters at the end of the line? It worked for me as i said up there in my last comment. |
|
@FTC55 ，Hey, thanks for your answer,but i had tried as you said,but it didnot work. Thanks for your advice! |
|
Which is what I asked you guys to do initially anyway
…On Tue, Dec 24, 2019, 9:43 PM He Leibin ***@***.***> wrote:
@FTC55 <https://github.com/FTC55> ，Hey, thanks for your answer,but i had
tried as you said,but it didnot work.
I then tried another version tensorflow implementation of CycleGAN,it
worked perfectly!
Thanks for your advice!
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#5?email_source=notifications&email_token=AIIR4WZQJTIG5MPXSUJZPODQ2LCGVA5CNFSM4HCG5XYKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEHTZPHI#issuecomment-568825757>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AIIR4WZ2MMGISXRNNYD7QDDQ2LCGVANCNFSM4HCG5XYA>
.
|
Hi !
At first, sorry for that I uploaded this problem as 'requests' at noon, I didn't notice the mistake when I wrote it.
I want to run this code, train it and see the results, but when I start training , the code will automatically quit after this:
OutOfRangeError (see above for traceback): RandomShuffleQueue '_2_shuffle_batch/random_shuffle_queue' is closed and has insufficient elements (requested 1, current size 0)
[[Node: shuffle_batch = QueueDequeueManyV2[component_types=[DT_FLOAT, DT_FLOAT], timeout_ms=-1, _device="/job:localhost/replica:0/task:0/device:CPU:0"](shuffle_batch/random_shuffle_queue, shuffle_batch/n)]]
How could this happen and how to do?
The text was updated successfully, but these errors were encountered: