nnet_train

aligner.multiprocessing.nnet_train(nnet_dir, egs_dir, mdl, i, num_jobs)[source]

Multiprocessing function that trains the neural net.

See:

for more details on the Kaldi binary this runs.

Also see https://github.com/kaldi-asr/kaldi/blob/master/egs/wsj/s5/steps/nnet2/train_pnorm_multisplice.sh for the original bash script that this function was based on.

Parameters:
nnet_dir : str

Directory of nnet training

egs_dir : str

Directory for training examples

mdl : str

Path to current model

i : int

Number of current iteration

num_jobs : int

The number of processes to use in calculation