You haven’t included a link to a lab, which you’ll need to do if you want someone to investigate this. In some labs that use Jupyter, it can take a bit of time for the lab to initialize, and you’ll see this before it is fully up.
Here it is:
Website i am trying to access: https://8888-port-5pmonxt4gu4siodv.labs.kodekloud.com/
Lab link: https://learn.kodekloud.com/user/playgrounds/playground-jupyter
And still getting bad gateway issue
I’m not seeing this. When I use the link supplied in the lab window, I start out with a window with a URL like:
https://8888-port-byen4ibeko6ljf2r.labs.kodekloud.com/lab/workspaces/auto-u
If I click the icon that has the label “Python 3 (ipython-3 kernel)”, it creates a window with a title similar to
https://8888-port-byen4ibeko6ljf2r.labs.kodekloud.com/lab/tree/Untitled.ipynb
It’s possible that you are dealing with networking issues, or an intermittent problem with the lab. Please try exiting the lab and starting it up again, to see if this recurs.
The reconnection error can have similar causes.
It’s not a network connectivity issue from my internet, it’s happening again and again for this specific lab.
It can be seen here
Could you summarize what you’re doing when this happens? I tried to install Pytorch, not using your technique, but by the one described here. It did take a while, and running top in another terminal, it looks like the system was CPU bound for some part of the process. But I can’t replicate your problem, still.
Working theory is that there’s some bottleneck in the lab, and you’re hitting it. Knowing better what you’re doing, I can see if I can hit it as well.
I am trying to follow and setup this demo on jupiter playground: PyTorch/section_4/demos/040-040-introduction-to-flask at main · kodekloudhub/PyTorch · GitHub
And as i don’t need codeserver here, i am just doing as follow
- Cloning the repo
- Activating conda
- pip install -r Pytorch/requirements.txt
I’m still not sure what’s going on at your end. I ran the the script, and it worked. I’m not sure how your course demo converts to our lab environment, but then, your PyTorch course does have labs; you may want to try them directly. TBH, I’m not sure why you’re having problems with the PG environment here.