Difference between: “ConvergenceWarning: Newton-cg failed to converge” and “the line search algorithm did not converge”

Question

Basically what the title says. I was running a multinomial logit with sklearn LogisticRegression and at one point it would throw the first error, so I