when I use this tool to train a model with own train.dict, here comes out the error of core dump,
but it did't dump in earlier steps. Anyone knows the reason?
and here is the code.
g2p-seq2seq --train dict_dir/mad.train --model_dir mymdl_dir/mandarin/
when I use this tool to train a model with own train.dict, here comes out the error of core dump, but it did't dump in earlier steps. Anyone knows the reason?
and here is the code. g2p-seq2seq --train dict_dir/mad.train --model_dir mymdl_dir/mandarin/
INFO:tensorflow:loss = 0.014552901, step = 5501 (76.460 sec) INFO:tensorflow:global_step/sec: 1.29732 INFO:tensorflow:loss = 0.03983097, step = 5601 (77.084 sec) INFO:tensorflow:global_step/sec: 1.29314 INFO:tensorflow:loss = 0.017567342, step = 5701 (77.329 sec) Segmentation fault (core dumped)