Pathways Language Model (PaLM): Scaling to 540B parameters

Re:Pathways Language Model (PaLM): Scaling to 540B parameters by yldedly on Wednesday April 06, 2022 @04:13AM (#30929706)

Take 100k joke and explanation tuples as training. Pattern match input joke to training jokes. Edit amalgamation of most structuraly similar training jokes to match content of input joke. Edit corresponding training explanation with same substitutions. Output explanation. Repeat with 100 input jokes, pick the best one.

[ Parent | on HN ]

Re:Pathways Language Model (PaLM): Scaling to 540B parameters by 1024core on Wednesday April 06, 2022 @11:37AM (#30933344)

But that is not what they did. Please read the paper.

[ Parent | on HN ]

Re:Pathways Language Model (PaLM): Scaling to 540B parameters by yldedly on Wednesday April 06, 2022 @12:51PM (#30934571)

I'm not saying that's what they did. I'm saying it's functionally equivalent.

[ Parent | on HN ]