bartowski

/

OLMo-2-1124-13B-Instruct-GGUF

like 12

Allen AI have updated that their original model was mistakenly mistrained.

They've released a fix with the same name, and the old mistrained version is now suffixed "-preview":

NOTE: 1/3/2025 UPDATE:

Upon the initial release of OLMo-2 models, we realized the post-trained models did not share the pre-tokenization logic that the base models use. As a result, we have trained new post-trained models. The new models are available under the same names as the original models, but we have made the old models available with a postfix "-preview".

So these are the preview quants.

Is it possible to re-generate quants for the new versions?

This would also apply to the 7B version: https://huggingface.co/bartowski/OLMo-2-1124-7B-Instruct-GGUF