Comment

Comment by npip99

In our case training and inferencing the models takes days, calculating all of the ELOs take 1min haha. So we didn't need to optimize the calculation.

But, we did need to work on numeric stability!

I have our calculations here: - https://hackmd.io/@-Gjw1zWMSH6lMPRlziQFEw/B15B4Rsleg

tldr; wikipedia iterates on <e^elo>, but that can go to zero or infinity. Iterating on <elo> stays between -4 and 4 in all of our observed pairwise matrices, so it's very well-bounded.

parent npip99 1 year ago

Discussion

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton