The Training API code for Tokenization is outdated in manual (2/2)

The code sample in the manual at <> in the section on Tokenizer training has is incorrect. The current code sample is:

try {
  model = TokenizerME.train("en", sampleStream, true, TrainingParameters.defaultParams());
} 

But TokenizerME.train() now has a new signature which requires a TokenizerFactory. The above does not compile with the 1.9.4 library version.