Google has the most big pre-trained models while Alibaba has the largest (openbmb.github.io)

2 comments

[ 2.6 ms ] story [ 11.7 ms ] thread

So the number of parameters metric here is the sum across all their models, right? There isn't yet a single model with 10 trillion parameters... right?

fishingboy 3y ago ↗

Yes, it is the sum of all models. But M6-10T (MoE) of Alibaba is a 10 trillion one.