Viewing profile — asaiacai

this list has a lot of false positives.

lmao

it's the perfect drug. You don't know how to code something up. Ask AI to implement it. It's broken? Ask AI to fix it for you. Will people become unable to fix things without it?

have you guys tested this against any resource/batch managers in k8s (i.e. kueue, volcano, apache yunikorn)? seems like this is a good fit for people who have already have a large …

its cool to see the iterative improvements to your model laid out, but for everything that workedm i imagine there were at least a million other things you also tried but didnt wor…

echoing this sentiment. I really do believe at some level there is strength/wisdom in being able to step away from a problem and return to it from a new perspective despite of what…

MFU is probably the best but requires application logic. You can export metrics at the infra level like SM efficiency. We explain it a bit how we used it to do some optimization. h…

John Ousterhout's "A Philosophy of Software Design" I liked. It was supposed to be assigned reading for Berkeley's data structures class CS61B, and I don't think I really internali…

power is also a good proxy. For example, we've had distributed runs that we monitored on WandB where one of our workers died in the middle and the rest were basically stalling on t…

totally agreed. A lot of our findings during this process is that there's still a lot of alpha in finding the right kernels for the job/model. We're hoping that in the future `torc…

I really hope this makes it easier to install/upgrade NVIDIA drivers on Linux. It's a nightmare to figure out version mismatches between drivers, utils, container-runtime...