[nullc]

Saying people lack the equipment to run them for inference isn't a good reason to not publish them. The astronomical training cost is a good reason to publish them.