glm 5.2 is already pretty good, soo...
post
replies:
Very likely the same architecture, just with more post training.
replies:
permalink
fedilink
source
parent
hideshow 1 child comment
What makes you say that? Not arguing, just curious.
replies:
permalink
fedilink
source
parent
Just repeating rumors (sorry, should have been clearer): GLM 5.3 is 5.2 with extra post training, their big upcoming model is 5.5. It kind of makes sense too; pretraining on a new architecture/size is expensive and it's natural for these companies to try to wring an extra minor version or two out of each one.


all 11 comments