Release code and/or models for Ensemble Debiasing on Hugging Face #1
Description
Hi @ruixilin 🤗
Niels here from the open-source team at Hugging Face. I discovered your work "Ensemble Debiasing Across Class and Sample Levels for Fairer Prompting Accuracy" through Arxiv and saw your paper is featured on https://huggingface.co/papers/2503.05157.
The paper page lets people discuss your work and find related artifacts. You can also claim the paper as yours, which will show up on your public profile, and link your GitHub repository.
I noticed in your GitHub README (https://github.com/NUS-HPC-AI-Lab/DCS) that your code is "Coming soon". That's fantastic news!
Once your code is ready, would you be interested in making it easily discoverable and usable on the Hugging Face Hub? We often see researchers hosting their code (e.g., as a custom model that can be loaded with from_pretrained via PyTorchModelHubMixin) directly on the Hub.
This enables better visibility and discoverability for your research. We can add relevant tags to make it easier for others to find, and link it directly from your paper page. If you plan to release any debiased model checkpoints (e.g., fine-tuned Llama-2 variants using your method) or new datasets, we'd also be thrilled to host them on the Hub.
If you're interested in hosting your code and/or any artifacts on the Hub once they are released, we'd be happy to provide guidance. You can find guides for uploading models here and datasets here.
Let me know if you'd like to discuss this further or need any assistance when your code is ready!
Kind regards,
Niels
ML Engineer @ HF 🤗