I need explanation, how the classifier is training (PHP-ML library)

Question

Can someone explain to me how the classifier is trained and finally predicted? I have in mind the last two lines of code below. I can't understand how it is working.

I'm writing a thesis about ML and text classification and it is last one of my problems.

My answer on it is: In this point the classifier is training. It is first of steps to create a model of answers. We are taking the samples and assign labels for them.

It is true?

$split = new StratifiedRandomSplit($dataset, 0.2, 42); 
$pipeline = new Pipeline([
    new TokenCountVectorizer($tokenizer = new NGramTokenizer(1, 3), new Polish()),
    new TfIdfTransformer()
], new SVC());

$pipeline->train($split->getTrainSamples(), $split->getTrainLabels());
$predicted = $pipeline->predict($split->getTestSamples());