How do I calculate the learning curve with the MLJ.jl library (Julia) using the number of input samples as parameter?

Julia's MLJ.jl provides a learning curve function to evaluate a machine learning model for various values of a parameter of choice. I would like to do this using the training sample size as the parameter to vary, just like Python's sklearn.model_selection.learning_curve. I've looked over MLJ's documentation but I couldn't find a solution: the problem is that the samples size is not a parameter of the model itself, which is what the range input argument of MLJ.learning_curve requires. Of course, writing a for loop that changes the sample size and evaluating the model every cycle is possible, but it seems to me that such a thing should be included in the library already (Python has exactly this function in sklearn, afterall).