yolov7 Model Version

#9
by cgz - opened

Hello,
I want to know which version of the model "yolov7" is this space using?
I got totally different prediction results between this space's "yolov7" and github's public model from the image below,
both detected items and prediction scores.

HuggingFace Space Result:
hugspace01.jpg

Public Model Result:
defaultmod01.jpg

Especially the blonde girl in right-middle behind, she never had been detected EXCEPT in this Hugging-Face space.
(here is the image source: https://s.yimg.jp/images/gyao/specials/2022/anime/summer/youkoso_2nd.jpg )
I have changed many versions of the yolov7 models, include the github's and the Hugging-Face hub's from the discussion, but neither of them ever got a result like this space's.
So I would like to know which version of the "yolov7" model is this space really using? (Not the yolov7-e6e or yolov7-e6)
And if it is possible for you to share it with me?

Thank you very much.

Sign up or log in to comment