Release WebXSkill on Hugging Face

[NielsRogge]

Hi @aiming-lab 🤗

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/2604.13318.

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, and add Github and project page URLs.

I saw on GitHub that the code and artifacts for WebXSkill are "coming soon." It'd be great to make the checkpoints (if any) and the datasets (such as the executable skill library or the synthetic trajectories) available on the 🤗 hub to improve their discoverability and visibility once you are ready to release them.

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 case you have custom model components or adapters, 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.

Uploading dataset

Would be awesome to make the skill library or trajectories 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 data in the browser.

Let me know if you're interested/need any guidance :)

Kind regards,

Niels
ML Engineer @ HF 🤗