i am facing issue in image classification with googlenet
Basically the shape math of the model layers needs to be checked so that the input feature size to the linear layer matches the defined weight shape. For example, this might be due to a flatten operation before the linear layer outputting a different shape than what is expected. However, since the model is pretrained you likely need to change the input shape (e.g., by resizing the image if it is an image input) to fix this.
ok let me try this i dont know will it work or not