BoredPigeon26 , are images from previous iterations still showing?
CostlyOstrich36
When I just started seeing this issue I could see images from previous iteration. Now some time passed and in the viewer I can't go back to the very first iterations.
BoredPigeon26 , you can scroll through iterations 🙂
at some point the button of "Older Images" become grey
Can you please provide a snippet of how the debug images are saved, also an example url would be useful :)
tf.summary.image('binary preds', plt_c1, max_outputs=5, step=self.get_step(),
description="Left to right: inputs | sigmoid prediction | labels (blue label, white ignore)") `plt_c1 is a tf tensor .
I have a local server not reachable from the internet.
BoredPigeon26 , when you copy and paste the link provided to you by the UI into the browser, can you see the image?
Not Found The requested URL was not found on the server. If you entered the URL manually please check your spelling and try again.
BoredPigeon26 , it looks like the file isn't accessible through your browser. Are you sure the remote machine files are accessible?
I killed and restarted the task (it created a new task, with same name, but different id) - and now I can see images again
CostlyOstrich36 , The issue does come back, after I kill, restart, and wait a few cycles.
Any luck to reproduce it?
Can you share an example of a working URL and a non-working URL?
Whenever I restart the task, I can see images for a while, then after few iteration, I can't anymore
If I right click and do copy image address - I get:
http://10.156.91.102:8081/semantics_net%252Ftraining/db_lms_preds_as_inputtest.86df109a78e94a21ae000bdd4e4cae97/metrics/Images/binary%20preds/Images_binary%20preds_00000041.jpeg?X-Amz-Date=1634799860337
The last one is working or not working?
ok CostlyOstrich36 , SuccessfulKoala55 :
The links differ in the very beginning -semantics_net%252Ftraining
becomes:semantics_net%2Ftraining
When I change it, I can see the image.
The project name is:"semantics_net/training"
There must be some unescaping for some reason
You say it happens at the middle of the experiment? any logs that you think might be related?
SuccessfulKoala55 No logs from clearml in the console, anywhere else I can look?
Nowhere I can think of right now... I'll try to poke around 🙂