ozanarkancan / char-ner Goto Github PK

Hi,

is it possible to share what are the library package requirements that you use to run this package?

i am pretty new to running people's codes. i have been trying to run your codes, but i cant run it successfully.

first there was issue with theano version unable to locate downsample, after which i downgraded to 0..
next, it produced ValueError: Cannot construct a ufunc with more than 32 operands (requested number were: inputs = 62 and outputs = 1) which through some google it shows it has to do with numpy version.
then i downgraded to 1.8.0 numpy and it produces ValueError: ('The following error happened while compiling the node', forall_inplace,cpu,scan_fn}(Elemwise{minimum,no_inplace}.0, Elemwise{sub,no_inplace}.0, Subtensor{int64:int64:int8}.0, Subtensor{int64:int64:int8}.0, IncSubtensor{InplaceSet;:int64:}.0, IncSubtensor{InplaceSet;:int64:}.0, Join.0, InplaceDimShuffle{x,0}.0, InplaceDimShuffle{x,0}.0, InplaceDimShuffle{x,0}.0), '\n', 'numpy.dtype has the wrong size, try recompiling')

Hopefully you can share with me the package requirements so i can run this successfully. Thanks!

I'm using Ubuntu 18.10 and installed the dependencies using pip3. When I run the example command:
python3 src/exper.py --activation bi-lstm --n_hidden 128 128 --drates .2 .5 .8 --lang cze

I get the following error:

WARNING (theano.tensor.blas): Using NumPy C-API based implementation for BLAS functions. Traceback (most recent call last): File "exper.py", line 12, in <module> from dataset import Dset File "/home/hasan/Documents/char-ner/src/dataset.py", line 4, in <module> import rep File "/home/hasan/Documents/char-ner/src/rep.py", line 59 print ' '.join(map(('{:^%d}'%mlen).format, [a[ci] for ci in cil] + [a[si]])) ^ SyntaxError: invalid syntax

Hi,
I added a Keras implementation of your great model:
https://github.com/0xnurl/keras_character_based_ner

Only the Viterbi decoder is missing for now.

Hi,
I found your paper interesting and I am trying to implement the same for english CoNLL 2003 dataset, is there any kind of preprocessing required for the data? Is it possible for you to provide the other languages datasets also which have been used for the paper.

I am getting the following error when trying to run the example command. What is going wrong here ?

$ python src/exper.py --activation bi-lstm --n_hidden 128 128 --drates .2 .5 .8 --lang cze
Traceback (most recent call last):
File "src/exper.py", line 238, in
main()
File "src/exper.py", line 221, in main
setup_logger(args)
File "src/exper.py", line 208, in setup_logger
ihandler = logging.FileHandler('{}/{}.info'.format(LOG_DIR,base_log_name), mode='w')
File "/home/sdn/anaconda3/envs/my_env_deep27/lib/python2.7/logging/init.py", line 913, in init
StreamHandler.init(self, self._open())
File "/home/sdn/anaconda3/envs/my_env_deep27/lib/python2.7/logging/init.py", line 943, in _open
stream = open(self.baseFilename, self.mode)
IOError: [Errno 2] No such file or directory: '/home/sdn/Desktop/char-ner/logs/onos-p4-dev-vm:cpu,activation:bi-lstm,n_hidden:128,128,fbmerge:concat,drates:0.2,0.5,0.8,recout:0,opt:adam,lr:0.001,norm:1,gclip:0,truncate:-1,n_batch:32,shuf:1,captrn:500,emb:0,lang:cze,fbias:0.0.info'