Lab 7: LOOCV for loop to fit polynomials from degree 1 to 5

In the line with `for i in range(1,6):` you have `model = lm.fit(X_current, y)` so you're training your model using the `X_test` data as well. Shouldn't the model be trained only with the training data? And then we would call `cross_val_score()` using the model and the `X` values?
