takuyahiraoka / dialogue-state-tracking-using-lstm Goto Github PK

rzai@rzai00:/prj/Dialogue-State-Tracking-using-LSTM/dstc4_traindev$ python main.py
/usr/local/lib/python2.7/dist-packages/fuzzywuzzy/fuzz.py:35: UserWarning: Using slow pure-python SequenceMatcher. Install python-Levenshtein to remove this warning
warnings.warn('Using slow pure-python SequenceMatcher. Install python-Levenshtein to remove this warning')
Using gpu device 0: GeForce GTX 1080 (CNMeM is disabled, cuDNN 5005)
Start to construct doc2vec from given dialogs.
['dstc4_train', 'dstc4_dev', 'dstc4_test']
<dataset_walker.dataset_walker object at 0x7f2d6796df90>
<dataset_walker.dataset_walker object at 0x7f2d6796dfd0>
<dataset_walker.dataset_walker object at 0x7f2db52f6f50>
Traceback (most recent call last):
File "main.py", line 32, in
LSTMWithBOWTracker.constructDoc2vec(nameDataS2V=["dstc4_train","dstc4_dev","dstc4_test"])
File "/home/rzai/prj/Dialogue-State-Tracking-using-LSTM/dstc4_traindev/scripts/LSTMWithBOW.py", line 658, in constructDoc2vec
for call in dataset:
File "/home/rzai/prj/Dialogue-State-Tracking-using-LSTM/dstc4_traindev/scripts/dataset_walker.py", line 42, in iter
raise RuntimeError,'Cant score : cant open labels file %s' % (labels_filename)
RuntimeError: Cant score : cant open labels file /home/rzai/prj/Dialogue-State-Tracking-using-LSTM/dstc4_traindev/data/21/label.json
rzai@rzai00:/prj/Dialogue-State-Tracking-using-LSTM/dstc4_traindev$

3493/3500(epoch/max epoch)
3494/3500(epoch/max epoch)
3495/3500(epoch/max epoch)
3496/3500(epoch/max epoch)
3497/3500(epoch/max epoch)
3498/3500(epoch/max epoch)
3499/3500(epoch/max epoch)
Doc2Vec was constructed with:
[<dataset_walker.dataset_walker object at 0x7efd47a65b10>, <dataset_walker.dataset_walker object at 0x7efd47a65550>]
, stored to LearnedDoc2Vec.d2v
Create o dictionary required in ensembler
In both of learning and tracking,
the tracker store output of LSTM at each turn in Sub.dial,
and tracker ignore the utterance which is not related to main task.
and tracker consider one subdialog at one sequence.
and the output of LSTM is combined with that of baseline.
and features made by M1s are used in input.
and distributed sentence reprentation is used instead of BOW and m/home/rzai/prj/Dialogue-State-Tracking-using-LSTM/dstc4_traindev/scripts/LSTMWithBOW.py:447: FutureWarning: comparison to None will result in an elementwise object comparison in the future.
if avrV==None:
eta info.
FuzzyMatchingResult.pic Dictionary are loaded.
M1Vec.pic.gzwas not found. Dictionary are newly created.
numClassFeature.pic was not found. learn() is required before tracking.
numSentenceFeature.pic was not found. learn() is required before tracking.
numM1Feature.pic was not found. learn() is required before tracking.
Doc2vec was loaded.
Start learning LSTM, and make dictionary file
Star make dictionary: variable name -> corresponding index of element in i/o vector
totalNumSlot:30
outputSize:5608
inputSize:2228
Start learning Network
Traceback (most recent call last):
File "main.py", line 44, in
lstm.learn(pathdataset=["dstc4_train"], EPOCHS_PER_CYCLE = 1, CYCLES = 1)#the pathdataset which same with that in latter ensembler should be set.
File "/home/rzai/prj/Dialogue-State-Tracking-using-LSTM/dstc4_traindev/scripts/LSTMWithBOW.py", line 347, in learn
if "frame_label" in label:
TypeError: argument of type 'NoneType' is not iterable
Write fuzzymatching result to file.
Write M1 vector comp. result to file.
7041,1 Bot

rzai@rzai00:~/prj$ sudo pip install pybrain
Downloading/unpacking pybrain
Downloading PyBrain-0.3.tar.gz (262kB): 262kB downloaded
Running setup.py (path:/tmp/pip_build_root/pybrain/setup.py) egg_info for package pybrain

Installing collected packages: pybrain
Running setup.py install for pybrain

Could not find .egg-info directory in install record for pybrain
Successfully installed pybrain
Cleaning up...
rzai@rzai00:/prj$ cd Dialogue-State-Tracking-using-LSTM/
rzai@rzai00:/prj/Dialogue-State-Tracking-using-LSTM$ cd dstc4_traindev/
rzai@rzai00:/prj/Dialogue-State-Tracking-using-LSTM/dstc4_traindev$ python main.py
Traceback (most recent call last):
File "main.py", line 16, in
from LSTMWithBOW import LSTMWithBOWTracker
File "/home/rzai/prj/Dialogue-State-Tracking-using-LSTM/dstc4_traindev/scripts/LSTMWithBOW.py", line 19, in
from pybrain.tools.customxml import NetworkWriter, NetworkReader
ImportError: No module named customxml
rzai@rzai00:/prj/Dialogue-State-Tracking-using-LSTM/dstc4_traindev$