When it is used as partition-by-any function to train the whole data sets as a single model, the function also supports LocalSGD, a variant of SGD, that uses LocalSGDIterations on each AMP to run multiple batch iterations locally followed by a global iteration.
The weights from all mappers are aggregated in a reduce phase and are used to compute the gradient and loss in the next iteration. LocalSGD lowers communication costs and can result in faster learning and convergence in fewer iterations, especially when there is a large cluster size and many features.
The function output is trained GLM models that is used as input to the TD_GLMPredict function. The output contains model statistics of MSE, Loglikelihood, AIC, and BIC. You can use TD_RegressionEvaluator,TD_ClassificationEvaluator, and TD_ROC functions to perform model evaluation as a post processing step.