I started using Pytorch. I find it very Nice, I tried to implement a basic linear regression model using a simple Feed forward neural network, I trained the model and I get to observe the loss decreasing over time but I didnt figure out how to get the accuracy of such a model with pytorch! in Keras that’s easy using
metrics = ["accuracy"] inside the compile function same in scikit learn it is also very easy using r2_score or other score functions. why such things didn’t exist in Pytorch? and if I have to implement an Accuracy function myself for a Linear regression model, how can I do that ?
ps: I know how to implement such a function for a Classification Problem because it have discrete values. In linear Regression it is different because we have continuous values that’s why I couldn’t find a way to implement an Accuracy function for it. I could just see the value of the loss and based on it say whether my model is good or not but it would be better if I achieve to implement an Accuracy function like in keras or sklearn to print the accuracy of the Model for every epoch. that would be great