Hi @ydeng9 🤗
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/2503.17352.
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 great to see that the OpenVLThinker-7B and OpenVLThinker-7B-v1.2 models are already on the 🤗 Hub and linked to the paper page!
I also noticed that the paper introduces "higher-quality SFT data" from your iterative self-improvement cycles, and the GitHub README links to a Hugging Face collection for this data (https://huggingface.co/collections/ydeng9/openvlthinker-v12-datasets-686f45e48d02e00b1585299e). I found several specific datasets within this collection, such as MathVista-OpenVLThinker-7B-SFT-v1.2, MMMUPRO-OpenVLThinker-7B-SFT-v1.2, EMMA-OpenVLThinker-7B-SFT-v1.2, and HallusionBench-OpenVLThinker-7B-SFT-v1.2.
It would be awesome to link these new datasets directly to your paper page as well, to improve their discoverability/visibility. We can add tags so that people find them when filtering https://huggingface.co/datasets and easily discover them from your paper page.
After uploaded (which they already are!), we can also link the datasets to the paper page (read here) so people can discover your work.
Let me know if you're interested or need any help regarding this!
Cheers,
Niels
ML Engineer @ HF 🤗
Hi @ydeng9 🤗
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/2503.17352.
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 great to see that the
OpenVLThinker-7BandOpenVLThinker-7B-v1.2models are already on the 🤗 Hub and linked to the paper page!I also noticed that the paper introduces "higher-quality SFT data" from your iterative self-improvement cycles, and the GitHub README links to a Hugging Face collection for this data (
https://huggingface.co/collections/ydeng9/openvlthinker-v12-datasets-686f45e48d02e00b1585299e). I found several specific datasets within this collection, such asMathVista-OpenVLThinker-7B-SFT-v1.2,MMMUPRO-OpenVLThinker-7B-SFT-v1.2,EMMA-OpenVLThinker-7B-SFT-v1.2, andHallusionBench-OpenVLThinker-7B-SFT-v1.2.It would be awesome to link these new datasets directly to your paper page as well, to improve their discoverability/visibility. We can add tags so that people find them when filtering https://huggingface.co/datasets and easily discover them from your paper page.
After uploaded (which they already are!), we can also link the datasets to the paper page (read here) so people can discover your work.
Let me know if you're interested or need any help regarding this!
Cheers,
Niels
ML Engineer @ HF 🤗