Error in running code of end to end learning program about qpth HOT 8 CLOSED

Hi @parimuns, sorry about the errors here from the older code here. This primal and dual residual look okay, I think the solution being returned here may be good and you can ignore the warning here and let the experiment keep running. How do the results look when you let it keep running?

\cc @priyald17

from qpth.

@bamos.Thanks for replying. I didn't get any solution. I ran the main.py file for testing in command interface with only 1 run and paramset (2) and then I got the above error. I didn't get any solution.
If there is something in code or folder which I am missing , please let me know.

I

from qpth.

Hmm, this message should just be a warning that the solver internally hit these NaNs and I think it will return the best solution it got so that the experiment can keep running

from qpth.

(But I've not run this code in a while and it may be erroring out because of this or some other reason)

from qpth.

Okay. Is there any new update in the code? Can you pls run this code and help me in this issue. This work is really beneficial to me and will be helpful in my research work
I am trying to contact @priyald17 , but she is not replying.
Thanks a lot.

from qpth.

Hi @parimuns, should be able to get back to you in the next few days!

from qpth.

@bamos feel free to resolve this issue here, as there's already an issue open on the e2e-learning repo.

from qpth.

Ah yeah, let's continue the discussion in locuslab/e2e-model-learning#8 if that's related here. Also the trace there isn't because of this qpth warning

from qpth.