These models are not of the same nature either. Their training was done in a different way. A uniform naming (even with explicit number of parameters) would still be misleading.