Congratulation on making it to top 100 trending models! 🎉
Hey PerceptronAI team,
I am Aritra from Hugging Face, and we are glad that this model has made it to the top 100 trending models of the day. Here are some ways you could increase the visibility of the model and make it more useful for the open source community.
- Creating a transformers based inference script would help people use the model quickly. I see that the
perceptronpackage is a requirement, but in the README of the repo it says that the model can also be run using Hugging Face Transformers. If that is the case addingtransformersas the libarary in the README.md would open up a lot of opportunities. - Demoing the model using a Hugging Face Space. That way people can try out the model and send quick feedback. Upgrading to Team would mean ZeroGPU access to the space.
- A blog post about the model, with Team subscription you could also publish a blog post around the model for a wider range of audience.
Hope this helps, and if you have any concerns do let me know.
(1) thanks to
@sergiopaniego
for bringing this in - we have some sample scripts in https://github.com/perceptron-ai-inc/perceptron/blob/main/huggingface/main.py using the transformers implementation
(2) Due to the visual nature (interactive bounding boxes) we are hosting the model on our UI to make it easier to interact with - we will update to include a link
(3) How can we get started and is there a way we can partner with huggingface on this?