Closed ftgreat closed 3 months ago
Thanks for your great paper. One question about aggregate evaluation, why metrics like MMLU are not included in aggregate evaluation?
We do MMLU evaluation for our Zamba model: https://huggingface.co/Zyphra/Zamba-7B-v1 .
For the dataset ablations we performed the same evaluations as Dolma and Refinedweb for comparison purposes.
Thanks for your great paper. One question about aggregate evaluation, why metrics like MMLU are not included in aggregate evaluation?