Improve discoverability of your work on HF

[NielsRogge]

Hi,

Niels here from the open-source team at Hugging Face. I discovered your work through the paper page: https://huggingface.co/papers/2407.15420 (feel free to claim the paper so that it appears at your HF account!).

However there are a couple of things which could improve the discoverability of your work, which I've listed below.

Add a model card

I see the model is already on the hub, but it's in a dataset repository: https://huggingface.co/datasets/hamacojr/LocoTrack-weights.

Would you be up for creating it in a model repository, rather than a dataset repository?

It would be great to also add a model card :) See here for some tips: https://huggingface.co/docs/hub/en/model-cards.

Moreover, the model could be linked with the paper, see here on how to do that: https://huggingface.co/docs/hub/en/model-cards#linking-a-paper

Uploading the PyTorch model

I see you're planning to release the PyTorch version as well. We recommend leveraging the PyTorchModelHubMixin to push your model to the hub and reload it using from_pretrained. This ensures a config.json along with safetensors weights are pushed to the hub.

Alternatively, we offer https://huggingface.co/docs/transformers/custom_models which ensures your model can be loaded using Transformers, with trust_remote_code=True.

Let me know if you need any help regarding this!

Cheers,

Niels
ML Engineer @ HF 🤗

[bhack]

And an HF demo of course.. 😉