I am doing external validation in negative binomial regression:

$$

Questions:

$(1)$ What are the steps to do Lasso $($L$1)$ and Ridge $($L$2)$ Regularization in SPSS$-29?$

$(2)$ In the same analysis, if AIC shows good model performance $($less than $10\%$ difference between training and testing data$),$ and Deviance and Pearson Chi$-$Square show over fitting $($over $10\%$ difference between training and testing data$),$ how can this be fixed?