compile_train_graphs

aligner.multiprocessing.compile_train_graphs(directory, lang_directory, split_directory, num_jobs, debug=False)

Multiprocessing function that compiles training graphs for utterances

See http://kaldi-asr.org/doc/compile-train-graphs_8cc.html for more details on the Kaldi binary this function calls.

Also see https://github.com/kaldi-asr/kaldi/blob/master/egs/wsj/s5/steps/train_mono.sh for the bash script that this function was extracted from.

Parameters:

directorystr

Directory of training (monophone, triphone, speaker-adapted triphone training directories)

lang_directorystr

Directory of the language model used

split_directorystr

Directory of training data split into the number of jobs

num_jobsint

The number of processes to use