Plans to release the model checkpoints on Hugging Face Hub?

[Vaibhavs10]

Hey hey! Congratulations on the brilliant release. I'm VB, and I'm working on the open-source team at Hugging Face.

I was wondering if you already plan to release the model checkpoints trained as part of the paper. The community can learn a lot from these checkpoints, evaluate, test, and even fine-tune them for specific tasks!

Happy to help with the model checkpoints release and the hub stuff if it helps.

Thanks again for releasing the codebase; I'm hoping y'all release the checkpoints, too.

Cheers, VB

[liuzechun]

Hi VB,

Thanks for reaching out!

Yes, that's in our plan. We are working with the legal team to make the models publicly available. It is taking longer than we expected. We'll update you here once it is approved by the legal team.

Thanks, Zechun

[Vaibhavs10]

Hi @liuzechun, That's brilliant! I'm looking forward to the checkpoints; there's a lot of community excitement about them!

Some quick doc links to help you when the legal approves them:

1. Our recommendation is to have one model checkpoint per repository. You can find all the details related to upload here in this guide: https://huggingface.co/docs/hub/models-uploading
2. You can use Collections to group and share all the model checkpoints together: https://huggingface.co/docs/hub/en/collections
3. In addition to this you can also add MobileLLM as a library on the hub so that all the models trained from it can be neatly organised: https://huggingface.co/docs/hub/en/models-adding-libraries#register-your-libraries-supported-tasks-on-the-hub

Let me know if you need any help with this! I'm at vaibhav [at] hf [dot] co or just DM me on Twitter (reach_vb)

[ahmadmustafaanis]

waiting for it

[redagavin]

Is there any progress?

[chenzx2]

waiting for it