Ottokar Siirak

Adding "-ac 2" solved a similar problem for me. This was suggested in a comment "That's not the same error at all?! Your new error is telling you that the...

I also had to change: `from tensorflow.contrib.cudnn_rnn import CudnnLSTM ` to: `from tensorflow.contrib.cudnn_rnn.python.ops.cudnn_rnn_ops import CudnnLSTM` in train_type.py line 25.

Punctuator uses Theano and it should support multiple cores to some extent http://deeplearning.net/software/theano/tutorial/multi_cores.html# , but I personally haven't experimented with training on CPU

You can also check if you have better luck using Tensorflow version: https://github.com/cadia-lvl/punctuation-prediction/tree/master/punctuator2tf2

Hi! it's probably the nltk tokenizer that is used in 2 scripts: demo_play_with_model.py and example/dont_run_me_run_the_other_script_instead.py To fix that you can modify the untokenizer (should work for both scripts): untokenizer =...

Hi! That's quite small dataset. I think I would split it into 80% training, 10% dev and 10% test data. The training file is obviously used for training the parameters...

You can find the solution here: https://github.com/ottokart/punctuator2/issues/17

Thanks, I'm glad you have found this project useful! I have not tried POS tag based model, but I would be very interested in the results if you happen to...

The model does look into the future. What you see here http://bark.phon.ioc.ee/punctuator/game is basically exactly what the model sees as well (Except the model sees numbers as NUM tokens and...

Hi, are you using the https://github.com/ottokart/punctuator2/releases/tag/v1.0 or the latest version?