Closed omarespejel closed 2 years ago
Hi @omarespejel - this looks really cool. I might be able to work on this after March due to several internal PhD deadlines. I'll set a reminder to potentially ping you if any questions then 😄
No worries! And yes, please let me know how I might be helpful 😃.
In Hugging Face we are really motivated to collaborate with and support the deep learning for healthcare initiative, and these models could be key 🤗. We would love to give them even more visibility.
This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.
Should this be reopened? :smile:
Should this be reopened? 😄
Hi @osanseviero, thanks for your consistent interest. Apologize for the late reply. Unfortunately, we currently do not have the bandwidth to maintain or integrate these models from DeepReg.
Meanwhile, if you are interested, the PyTorch-based code base and models of our recent work (paper 1, preprint 2) on medical imaging are available and still under active maintenance by @kate-sann5100. Please feel free to open an issue there 😄
Hi DeepReg team!
Your pre-trained models seem amazing! I see you host and share the DeepReg demo models on your own repo. Would you be interested in sharing your models in the Hugging Face Hub?
This integration would allow you to freely download/upload models, and make your work more accessible and visible to the rest of the ML community. We can help you set up a DeepReg organization (examples, Facebook AI y Stanford NLP).
Creating the repos and adding new models should be a relatively straightforward process. This is a step-by-step guide explaining the process in case you're interested. Please let us know if you would be interested and if you have any questions.
Some of the benefits of sharing your models through the Hub would be:
- Presence in the HF Hub might lower the entry of barrier to DeepReg models as well as increase its visibility.
Additionally, we have a library to programmatically access repositories (both downloading pretrained models and pushing, with a lot of nice things such as filtering, caching, etc). If we want to try out this integration, I would suggest you add one or two models manually and then use the
huggingface_hub
library to implement downloading those models programmatically from the DeepReg demo models. You might want to check our documentation to read more about it.Relevant references:
Happy to hear your thoughts,
Omar and the Hugging Face team
cc @osanseviero @abidlabs (@NMontanaBrown thanks for the reference!)