Hi @hukcc 🤗
Niels here from the open-source team at Hugging Face. I discovered your work on Arxiv and noticed your paper, "D-CoDe: Scaling Image-Pretrained VLMs to Video via Dynamic Compression and Question Decomposition," which introduces an efficient training-free framework for Video LLMs. Your paper is already featured on our paper page: https://huggingface.co/papers/2510.08818.
I saw on your GitHub repository (https://github.com/hukcc/D-CoDe) that the code for D-CoDe will be released soon. It'd be great to make the code for your D-CoDe framework available on the 🤗 Hub once it's ready, to improve its discoverability and visibility within the AI community.
The Hugging Face Hub (huggingface.co) is a great platform for researchers to share their work, including code, models, datasets, and Gradio demos. You can also claim your paper on its HF page, which will show up on your public profile, and add GitHub and project page URLs.
We can help you set up a repository for your code on Hugging Face, or even assist in creating a Hugging Face Space for a live demo of D-CoDe. We can also provide a ZeroGPU grant, which gives you access to A100 GPUs for free for such a demo.
Let me know if you're interested or need any guidance once the code is ready for release!
Cheers,
Niels
ML Engineer @ HF 🤗
Hi @hukcc 🤗
Niels here from the open-source team at Hugging Face. I discovered your work on Arxiv and noticed your paper, "D-CoDe: Scaling Image-Pretrained VLMs to Video via Dynamic Compression and Question Decomposition," which introduces an efficient training-free framework for Video LLMs. Your paper is already featured on our paper page: https://huggingface.co/papers/2510.08818.
I saw on your GitHub repository (https://github.com/hukcc/D-CoDe) that the code for D-CoDe will be released soon. It'd be great to make the code for your D-CoDe framework available on the 🤗 Hub once it's ready, to improve its discoverability and visibility within the AI community.
The Hugging Face Hub (huggingface.co) is a great platform for researchers to share their work, including code, models, datasets, and Gradio demos. You can also claim your paper on its HF page, which will show up on your public profile, and add GitHub and project page URLs.
We can help you set up a repository for your code on Hugging Face, or even assist in creating a Hugging Face Space for a live demo of D-CoDe. We can also provide a ZeroGPU grant, which gives you access to A100 GPUs for free for such a demo.
Let me know if you're interested or need any guidance once the code is ready for release!
Cheers,
Niels
ML Engineer @ HF 🤗