gdeotale/cifar_gradcam_visualization

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

App link

https://huggingface.co/spaces/gdeotale/cifargradcamvisualization

App configuration

1. 1.Upload image from ('plane', 'car', 'bird', 'cat', 'deer', 'dog', 'frog', 'horse', 'ship', 'truck') these classes for better results.
2. 2.Adjust slider1 for choosing layer from which u want to see gradcam output.
3. 3.Adjust slider2 for choosing top n classes predicted by model.
4. 4.choose show gradcam image checkbox to visualise gradcam results
5. 5.choose show misclassified image to visualize 10 misclassified image
6. 6.submit

Model

https://huggingface.co/spaces/gdeotale/cifargradcamvisualization/blob/main/final_dict.pth

lightning cifar model

https://huggingface.co/spaces/gdeotale/cifargradcamvisualization/blob/main/Model/Litcifarmodule.py is the wrapper model that needs to be called before loading model https://huggingface.co/spaces/gdeotale/cifargradcamvisualization/blob/main/Model/custom_resnet.py is the base model taht is called from above model

Model accuracy