MPT-7b-instruct was not found on hub

#31
by JeromeMore - opened

Hi,

I would like to submit mpt-7b-instruct to eval queue but got the error : "MPT-7b-instruct was not found on hub!". However the model is present mpt-7b-instruct model card.

image.png

Regards,

Did they add the models to transformers? I wonder if that's the issue. We'll look!

Open LLM Leaderboard org

The mosaic/mpt models can execute abritary code that requires trust_remote_code to be launched. The eval queue does not support these models due to the inherent risk of loading them on our cluster. We made a special exception for mosaic/mpt-7b at the author's request.

Maybe similar case for THUDM/chatglm-6b?

JeromeMore changed discussion status to closed

Sign up or log in to comment