Speeding Up Open-Source AI Research | Illia Polosukhin
Nothing is loaded from YouTube until you press play, and then without tracking cookies.
About this video
llia Polosukhin from NEAR discusses the challenges of accelerating open-source AI research to keep up with rapid industry advancements. He shares insights on coordinating research, fostering innovation, and how faster experimentation can lead to discoveries. #OpenSourceInnovation #TechResearch #Collaboration #InnovationSolutions #SoftwareDevelopment #ResearchChallenges #ModelArchitecture #DatasetRefinement #Experimentation #TechTalks
Transcript
161 wordsCaptions as YouTube publishes them, usually machine-generated, so product and people names can be misheard.
- 0:00
being in one in this research environments where kind of right you know we see they're making really rapid progress compared to open source it's just like such a you know kind of stranded environment so that's kind of where we've been spending more time is like how do we kind of how do we coordinate research more effectively in open source in such a way that like when there is enough novel kind of developments both in model architecture
- 0:27
and training and data set refinement in in other aspects you can then go and train like one of this larger models uh but first you kind of need to have this Innovations and you know as as can of was research it's a lot of experimentation a lot of failure and so how do we how do we increase the speed of failure pretty much so that we can come toward you know golden nuggets of of innovation from this