Release WorldPM 15M Preference Data and link models on Hugging Face #4
Description
Hi @refrain-wbh 🤗
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/2505.10527.
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's fantastic to see that you've already released the WorldPM model checkpoints (e.g., Qwen/WorldPM-72B) on the Hugging Face Hub! To further improve their discoverability and visibility, we would love to help you link these models to your paper page on Hugging Face. You can read more about linking models to a paper here: https://huggingface.co/docs/hub/en/model-cards#linking-a-paper.
I also noticed in your paper that you collected a "15M-scale data" from public forums to train WorldPM. Would you be interested in making this valuable dataset available on the 🤗 Hub as well? Hosting it on Hugging Face would greatly enhance its visibility, enable better discoverability, and allow the community to easily use it with:
from datasets import load_dataset
dataset = load_dataset("your-hf-org-or-username/your-dataset")You can find a guide for uploading datasets here: https://huggingface.co/docs/datasets/loading.
Additionally, the dataset viewer allows people to quickly explore the first few rows of the data directly in the browser.
Let me know if you're interested or need any help regarding this!
Cheers,
Niels
ML Engineer @ HF 🤗