TIMIT_MLP_fmllr.cfg

[data]
out_folder=TIMIT_MLP_fmllr

tr_fea_scp=/home/mirco/pytorch_MLP_for_ASR/fmllr_lists/train_split.000,/home/mirco/pytorch_MLP_for_ASR/fmllr_lists/train_split.001,/home/mirco/pytorch_MLP_for_ASR/fmllr_lists/train_split.002,/home/mirco/pytorch_MLP_for_ASR/fmllr_lists/train_split.003,/home/mirco/pytorch_MLP_for_ASR/fmllr_lists/train_split.004
tr_fea_opts=
tr_lab_folder=/home/mirco/kaldi-trunk/egs/timit/s5/exp/dnn4_pretrain-dbn_dnn_ali
tr_lab_opts=ali-to-pdf


dev_fea_scp=/home/mirco/pytorch_MLP_for_ASR/fmllr_lists/dev_split.000
dev_fea_opts=
dev_lab_folder='/home/mirco/kaldi-trunk/egs/timit/s5/exp/dnn4_pretrain-dbn_dnn_ali_dev'
dev_lab_opts='ali-to-pdf'

te_fea_scp=/home/mirco/pytorch_MLP_for_ASR/fmllr_lists/test_split.000
te_fea_opts=
te_lab_folder=/home/mirco/kaldi-trunk/egs/timit/s5/exp/dnn4_pretrain-dbn_dnn_ali_test
te_lab_opts=ali-to-pdf

count_file=/home/mirco/kaldi-trunk/egs/timit/s5/exp/dnn4_pretrain-dbn_dnn/ali_train_pdf.counts

[architecture]
hidden_dim=1024
N_hid=4
drop_rate=0.2
use_batchnorm=1
cw_left=5
cw_right=5
seed=1234
use_cuda=1

[optimization]
N_ep=24
lr=0.08
halving_factor=0.5
improvement_threshold = 0.001
batch_size=128
save_gpumem=0