Hi @ashimdahal 🤗
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/2510.01009.
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.
I saw that you've already shared the preprocessed frames for ReasonVQA on the hub (ashimdahal/povqa-preprocessed), which is great for the community!
It'd be awesome to also make the pre-trained checkpoints and the full dataset annotations available on the 🤗 hub, to improve their discoverability and visibility.
Uploading models
Would you like to host the pre-trained QLoRA adapters (the SFT and DPO policies) on https://huggingface.co/models?
Since you are using peft, you can easily use the push_to_hub method. We encourage researchers to push each adapter/checkpoint to a separate model repository, so that download stats and model cards work effectively. We can then also link these checkpoints directly to the paper page.
Uploading dataset
Would it also be possible to upload the final ReasonVQA annotations (the QA + rationale triplets currently in the zip file) directly to Hugging Face? This would allow people to load the entire dataset via:
from datasets import load_dataset
dataset = load_dataset("ashimdahal/ReasonVQA")Besides that, there's the dataset viewer which allows people to quickly explore the rationales and QA pairs directly in the browser.
Let me know if you're interested or need any help regarding this!
Cheers,
Niels
ML Engineer @ HF 🤗
Hi @ashimdahal 🤗
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/2510.01009.
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.
I saw that you've already shared the preprocessed frames for ReasonVQA on the hub (ashimdahal/povqa-preprocessed), which is great for the community!
It'd be awesome to also make the pre-trained checkpoints and the full dataset annotations available on the 🤗 hub, to improve their discoverability and visibility.
Uploading models
Would you like to host the pre-trained QLoRA adapters (the SFT and DPO policies) on https://huggingface.co/models?
Since you are using
peft, you can easily use thepush_to_hubmethod. We encourage researchers to push each adapter/checkpoint to a separate model repository, so that download stats and model cards work effectively. We can then also link these checkpoints directly to the paper page.Uploading dataset
Would it also be possible to upload the final ReasonVQA annotations (the QA + rationale triplets currently in the zip file) directly to Hugging Face? This would allow people to load the entire dataset via:
Besides that, there's the dataset viewer which allows people to quickly explore the rationales and QA pairs directly in the browser.
Let me know if you're interested or need any help regarding this!
Cheers,
Niels
ML Engineer @ HF 🤗