Keep on to blur preview images; turn off to show them clearly
now i need to spend some more time and attune myself to the writing ability or lack of for this new model. also calibrate how I feel about it's intelligence inc or decrease.
RL and distributed training • eXperiments lab
if tile-based primitives generalize across nvidia, apple and amd, then a unified, high perf programming model for multiple accelerators is plausible. that weakens the “cuda moat” and opens the ai compute landscape. https://t.co/cT1W7yUwDx
RL and distributed training • eXperiments lab
you do not need to see all the things that came out today. you just need to see THE ONE thing that came out today
if tile-based primitives generalize across nvidia, apple and amd, then a unified, high perf programming model for multiple accelerators is plausible. that weakens the “cuda moat” and opens the ai compute landscape. https://t.co/cT1W7yUwDx
RT @SzymonOzog_: I made multi-node inference 25% faster with a low latency allreduce. 🧵 My first post stated: "A goal would be to be able t…
RL and distributed training • eXperiments lab
RT @natolambert: I’m starting a new series of interviews on @interconnectsai with all the leading open model labs around the world to show…
RL and distributed training • eXperiments lab
awarded honorary points to myself for making this part of the noosphere. you're welcome.
RL and distributed training • eXperiments lab