Geoffrey Negiar
Geoffrey Negiar
It would also be really nice to benchmark algorithms on nonconvex problems. Do the usual practical acceleration methods (backtracking line search) work here?
I meant does it work for accelerating convergence in practice on non-convex problems?
For now, the only parallelization that's done in this part of the code base is for [sampling the batches](https://github.com/openopt/copt/blob/e204ec04dfda61b6938ac2af7dcc70b9bd3e9b10/copt/utils.py#L83), once per epoch. Is it possible that the 100% on the...
Does that mean that it would be nice in practice ? Since for applications, you won't be computing the gap very often.
None that I've found. Very probably this is just applying momentum in the right place(s). :) The setting I'm looking at is \min_{X\in C, Y} f(X + Y) + g(Y),...
Thanks for the refs! On VR for Neural Nets: https://arxiv.org/abs/1812.04529
TODO: - [x] Stochastic Robust PCA works! - [x] Get the MNIST example to work
This is nice, thanks a lot @ThomVett! Do you think you can check out #55 for more github action fun? It seems like public repos have unlimited gitub action time,...
Right now, the docs are built with `sphinx` by running `make html` from within the `doc` folder. Then running `make upload` uploads it to our server and refreshes the website....
@fabianp I don't have the permissions to access the secret keys to set up the github action for uploading to our server. Could you please add the two secrets required...