For a Custom Model, you write Python code. Your Python code must create a 'clf' variable.
This clf must be a scikit-learn compatible model, ie, have at least fit(X,y) and predict(X) methods, and follow the documentation on handling attributes.

More information available in the documentation.

Warning: you have selected the variable {{mlTaskDesign.weight.sampleWeightVariable}} as a weight variable.
Please make sure that the 'fit' method of the class of your custom 'clf' accepts 'sample_weight' as an argument.