Add OLMO2 model from AI2

[fpaupier]

Given the tech report published on Dec 31 2024 available here https://arxiv.org/abs/2501.00656
The OLMO 2 team did a great job of documenting several steps for their pretraining, instruction tuning, datasets used and evals method. I'd love to see how they perform in "openness score" against the other models currently listed.

[mdingemanse]

Thanks for the heads-up @fpaupier! Stay tuned as we work on a major next release of our openness framework — we've noted this one for inclusion