Acoustic model for Khalka Mongolian, trained on a corpus of 14 speakers (approx. 2 hours of speech) in Kaldi. Model and example output of forced alignment using the Montreal Forced Aligner can be found here.