I'm familiar with the literature. I do not like the research. It is bad, it's rarely reproducible, source code is almost never public thus time/convergence mea

I'm familiar with the literature. I do not like the research. It is bad, it's rarely reproducible, source code is almost never public thus time/convergence measurements are meaningless.