Talkie: a 13B vintage language model from 1930talkie-lm.com773 points·by jekude·il y a 3 mois·331 comments
jekude·il y a 25 jours·discussThe abstract implies that strictly local updates are a hinderance to high res, however i would have thought there would be an interesting way to get speed up gains from neighbor-only traffic on GPUs CAM-style. am i making that up?
jekude·le mois dernier·discussstarted doing this but life got in the way, would love to pick it back up at some pointhttps://github.com/planetlambert/turing
jekude·il y a 3 mois·discussWhat would physical EML gates be implemented in reality?Posts like these are the reason i check HN every day
jekude·il y a 5 mois·discussif the entire open web is vulnerable to being sybil attacked, are we going to have to take this all underground?
jekude·il y a 5 mois·discussHow does this compare to https://encore.dev/ or other infrastructure-inside-application-code solutions?
jekude·il y a 9 mois·discussI have been asking myself for a while: Why has Artificial Life not had its AlexNet moment with GPUs yet?