Nvidia is breaking the trend of sesame street themed models with the new MegatronLM.

bad dl joke 

Hopefully it's good at approximating the global optimus prime.

Show thread

But also 8.3 billion parameters? Seriously? I can't but help feel like this is designed to sell Tesla GPUs. That's gonna take a lot of VRAM.

Show thread
Sign in to participate in the conversation
Scholar Social

Scholar Social is a microblogging platform for researchers, grad students, librarians, archivists, undergrads, academically inclined high schoolers, educators of all levels, journal editors, research assistants, professors, administrators—anyone involved in academia who is willing to engage with others respectfully.