Skip to content

Release artifacts (post-trained models and datasets) on Hugging Face #1

Open

Description

@NielsRogge

Hi @YinyiLuo 🤗

Niels here from the open-source team at Hugging Face. I discovered your work through Hugging Face's daily papers as yours got featured: https://huggingface.co/papers/2604.10784.
The paper page lets people discuss about your paper and lets them find artifacts about it (your models, datasets or demo for instance), you can also claim
the paper as yours which will show up on your public profile at HF, add Github and project page URLs.

It'd be great to make the checkpoints (such as the 14B Bagel variant and the post-trained weights like Bagel + RecA, UniCot, and UniGame) and the novel datasets you introduced (like Uni-MMMU and ImgEdit-Bench) available on the 🤗 hub, to improve their discoverability/visibility and facilitate the reproducible comparisons you highlight.

We can add tags so that people find them when filtering https://huggingface.co/models and https://huggingface.co/datasets.

Uploading models

See here for a guide: https://huggingface.co/docs/hub/models-uploading.

In this case, we could leverage the PyTorchModelHubMixin class which adds from_pretrained and push_to_hub to any custom nn.Module. Alternatively, one can leverages the hf_hub_download one-liner to download a checkpoint from the hub.

We encourage researchers to push each model checkpoint to a separate model repository, so that things like download stats also work. We can then also link the checkpoints to the paper page.

Uploading dataset

Would be awesome to make the datasets available on 🤗 , so that people can do:

from datasets import load_dataset

dataset = load_dataset("your-hf-org-or-username/your-dataset")

See here for a guide: https://huggingface.co/docs/datasets/loading.

Besides that, there's the dataset viewer which allows people to quickly explore the first few rows of the data in the browser.

Let me know if you're interested/need any help regarding this!

Cheers,

Niels
ML Engineer @ HF 🤗

Activity

  1. YinyiLuo commented on Apr 14, 2026

    @YinyiLuo
    Collaborator

    Hi, thanks a lot for reaching out!

    We’ve updated the repository to include the post-training weights.

  2. NielsRogge commented on Apr 14, 2026

    @NielsRogge
    Author

    Hey @YinyiLuo, it's great to see that you've uploaded the post-training weights to the Hugging Face hub!

    However, it seems they are not yet linked to the paper page. Could you link them by mentioning the paper URL (preferably https://huggingface.co/papers/2604.10784, otherwise the arXiv abstract or PDF URL) in the README of the Hugging Face model repo?

    This will make the artifact automatically show up on the paper page.

    For example, you could add a reference to the TorchUMM paper in the model card(s) under aifronter/post_train, such as adding a line like:

    This model is introduced in the paper TorchUMM: A Unified Multimodal Model Codebase for Evaluation, Analysis, and Post-training.

    See also here for more information: https://huggingface.co/docs/hub/model-release-checklist#write-a-comprehensive-model-card

    Thanks!

  3. NielsRogge commented on Nov 14, 2026

    @NielsRogge
    Author

    Hi again,

    Separately, I've also made the paper and 5 paper-native evaluations available on Papers with Code.

    The paper has results on Image Understanding and Image generation task pages.

    Would it be possible to verify these results and let me know if any score, model name, benchmark protocol, or openness metadata should be corrected?

    You can also edit the task, methods, project page, and GitHub URL directly from the paper page using your Hugging Face account.

    Kind regards,

    Niels

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions