def init_params_shared(params):
mylog('Start Initializing Shared Parameters')
params_shared = OrderedDict()
for key,value in params.iteritems():
params_shared[key] = theano.shared(params[key], name = key)
mylog('Stop Initializing Shared Parameters')
return params_shared
writeFile(prefix + '.reference', reference)
writeFile(prefix + '.summary', summary)
writeFile(prefix + '.counts', time_data)
def loadFromText(fName):
f = codecs.open(fName, 'r', encoding='utf-8')
result = []
for l in f:
line = l.strip().split()
result.append(line)
return result
if __name__ == '__main__':
log = mylog()
dataoptions = optionsLoader(log, True)
# Load the Vocabulary and Features and Dataset First
Vocab_Giga = loadFromPKL('giga_new.Vocab')
Vocab = {
'w2i': Vocab_Giga.w2i,
'i2w': Vocab_Giga.i2w,
'i2e': Vocab_Giga.i2e
}
Features_Giga = loadFromPKL('features.Embedding')
I2Es = []
for feat in dataoptions["featList"]:
I2Es.append(Features_Giga[feat].i2e)
dataset = data_loader(Vocab, dataoptions, log)
import codecs, sys
from mylog.mylog import mylog
from vocabulary.vocabulary import Vocabulary, I2E
from utility.utility import *
from options_loader import *
if __name__ == '__main__':
#Vocab = loadFromPKL('../../vocab/gigaword.pkl')
log_vocab = mylog(logFile='log/log_vocab')
options = optionsLoader(log_vocab, True)
fileName = options['primary_dir']
inputCorpus = [fileName + options['trainSet'] + '.Ndocument']
outputCorpus = [fileName + options['trainSet'] + '.Nsummary']
Vocab = Vocabulary(options,
inputCorpus=inputCorpus,
outputCorpus=outputCorpus)
log_vocab.log(
str(Vocab.full_size) + ', ' + str(Vocab.n_in) + ', ' +
str(Vocab.n_out))
saveToPKL(fileName + sys.argv[1] + '.Vocab', Vocab)
f = codecs.open(fileName + sys.argv[1] + '.i2w', 'w', encoding='utf-8')
for item in Vocab.i2w: