I'd wish people stop using "open sourcing" when speaking about models.
Open sourcing is about being able to change and replicate builds, they make the models "freely available" but the recipe on how they are made is kept secret.
It's akin to being able to download Windows shareware executables and calling that "open source" when nothing related to how the executables are build is available.
Even if they did release the code, that would not help you much unless you have the $ for traning and the talented individuals for pipelining and distributed training.
I know they won't, but a man can dream.