printing model configuration of lazypredict models

Question

I am using lazypredict to find the most suitable regression model. For my dataset, it is giving SVR as most suitable model (as per the RMSE = 0.02 and R square value = 0.85). However, when I use given SVR regression model, I am getting different error indices (RMSE = 0.13 and R square =0.35). Even after trying different values for model parameters, I could not get the RMSE and R square values given by lazypredict. I would like to know how to get the model parameters values on basis of which lazypredict gives the list of models and corresponding error values. Here is my code.

xt = pd.read_csv('data.csv') 
xx = xt.iloc[:, [5,7,8,10,11]].values 
yy = xt.iloc[:,4].values 

xtrain, xtest, ytrain, ytest=train_test_split(xx,yy,test_size=0.25,random_state=0)
reg = LazyRegressor(predictions=True)
models, predictions = reg.fit(xtrain, xtest, ytrain, ytest)
print(models)

Adjusted R-Squared  R-Squared  RMSE  Time Taken
Model                                                                         
NuSVR                                        0.40       0.43  0.18        0.05
SVR                                          0.40       0.43  0.18        0.02
MLPRegressor                                 0.37       0.40  0.18        0.25

I want to know what values/arguments are being put by lazypredict for the regression models regressor_nu = NuSVR(kernel=?, nu=?, C=?)