Release LoopSpeech model checkpoints and AnchorSpeech dataset on Hugging Face

[NielsRogge]

Hi @benyou 🤗

I'm Niels and work as part of the open-source team at Hugging Face. I discovered your work through Hugging Face Daily Papers: https://huggingface.co/papers/2609.05592

The paper page lets people discuss your paper and find related artifacts such as models, datasets, and demos. You can also claim the paper as yours so it shows up on your public profile on Hugging Face, and add GitHub/project page URLs.

I saw in your GitHub README that code, models, data, and the public paper link will be added as they become available, and that the roadmap includes model checkpoints and AnchorSpeech data. When those are ready, it would be great to host the model checkpoints and the AnchorSpeech dataset on the 🤗 hub to improve their visibility and discoverability. We can add metadata tags so people can find them when filtering https://huggingface.co/models and https://huggingface.co/datasets, and we can link them to the paper page.

Uploading models

See here for a guide: https://huggingface.co/docs/hub/models-uploading.

For custom PyTorch models, you can use the PyTorchModelHubMixin class, which adds from_pretrained and push_to_hub. Alternatively, people can download checkpoints with hf_hub_download.

We encourage researchers to push each model checkpoint to a separate model repository so download stats and linking work cleanly.

Uploading dataset

It would be great to make AnchorSpeech available on Hugging Face so people can do:

from datasets import load_dataset

dataset = load_dataset("your-hf-org-or-username/AnchorSpeech")

See here for a guide: https://huggingface.co/docs/datasets/loading.

There is also the dataset viewer, which lets people quickly explore the first few rows in the browser.

After uploading, we can also link the models and dataset to the paper page so people can discover your work.

Let me know if you're interested or need any guidance.

Kind regards,

Niels
ML Engineer @ HF 🤗