HoG features taking longer time in Google CO-LAB

I have been doing this for 2 days now.

I want to make a model using sklearn that uses the HoG (Histogram oriented Gradients) features to classify the image (MNIST dataset having 70,000 images both training and testing combine) in Google CO-LAB, but somehow the model is taking much time then it is intended, and I don't know what mistake I am making. Can anyone help me out?

My HoG feature vector size is 6084. I have also divided the training data into train and test sets. I also used both GPU and TPU for the speed enhancement but no positive result has been found.