Description of the feature

Currently, it seems like there are models evaluated on MMTEB v1, and while that is true, most models are not evaluated on all tasks.

This make it seem like the benchmark is more complete than it is. I would instead display only fully evaluted models

We could add an info box for models in the overview that states "A total of {n} models is evaluated on all tasks. A total of {N} models is evaluated on at least one task within the benchmark."