Pytorch Model deployment using Windows Powershell with hugging face to commit changes "RuntimeError: Could not infer dtype of numpy.uint8" error and "TypeError: ConnectionPool.init() got an unexpected keyword argument 'socket_options'"

[M12feiz]

Following are the steps i followed that worked on May 5 2025:

1. download the archive from colab (foodvision_mini.zip)

2. create a folder on desktop (foodvisiondemo) using: "cd Desktop" followed by mkdir foodvisiondemo

3. go into this directory cd foodvisiondemo

4. create hugging face space and get its cloning link git clone https://huggingface.co/spaces/username/space_name

5. paste clonning link in terminal: C:\Users\Pc\Desktop\foodvisiondemo> git clone https://huggingface.co/spaces/username/space_name

6. change directory: Desktop\foodvisiondemo> cd foodvision_mini123

7. extract zip file ito this directory using command in format: Desktop\foodvisiondemo\space_name> Expand-Archive -Path "C:\Users\Pc\Downloads\foodvision_mini.zip" -DestinationPath "C:\Users\Pc\Desktop\foodvisiondemo\space_name"

8. initialize git in current working directory: git init
   now run the following commands in sequence:

9. git remote -v

10. git lfs install

11. git lfs track "09_pretrained_effnetb2_feature_extractor_pizza_steak_sushi_20_percent.pth"

12. git lfs track ".pth"

13. git add .gitattributes

14. git config --global core.autocrlf false

15. git add *

16. git status

17. git commit -m "first commit"

18. git push

19. there will be 2 errors while building: "RuntimeError: Could not infer dtype of numpy.uint8" error and "TypeError: ConnectionPool.init() got an unexpected keyword argument 'socket_options'" To resolve these, change versions in requirement.txt by editing file in huggingface files and adding following:
    gradio==2.6.4
    numpy==1.24.4

Your Application will start running