def train(training_list, model_path, format, use_lstm, logfile=None):
# Read the data into a Document object
docs = []
for txt, con in training_list:
doc_tmp = Document(txt, con)
docs.append(doc_tmp)
# file names
if not docs:
print 'Error: Cannot train on 0 files. Terminating train.'
return 1
# Create a Machine Learning model
model = ClinerModel(use_lstm)
# Train the model using the Documents's data
model.train(docs)
# Pickle dump
print '\nserializing model to %s\n' % model_path
with open(model_path, "wb") as m_file:
pickle.dump(model, m_file)
model.log(logfile, model_file=model_path)
model.log(sys.stdout, model_file=model_path)
def train(training_list, model_path, format, logfile=None):
# Read the data into a Document object
docs = []
for txt, con in training_list:
#try:
doc_tmp = Document(txt, con)
docs.append(doc_tmp)
#except Exception, e:
# exit( '\n\tWARNING: Document Exception - %s\n\n' % str(e) )
# file names
if not docs:
print 'Error: Cannot train on 0 files. Terminating train.'
exit(1)
# Create a Machine Learning model
model = ClinerModel()
# Train the model using the Document's data
model.fit_from_documents(docs)
# Pickle dump
print '\nserializing model to %s\n' % model_path
with open(model_path, 'wb') as f:
pickle.dump(model, f)
model.log(logfile , model_file=model_path)
model.log(sys.stdout, model_file=model_path)
def train(training_list,
model_path,
format,
use_lstm,
logfile=None,
val=[],
test=[]):
# Read the data into a Document object
train_docs = []
for txt, con in training_list:
doc_tmp = Document(txt, con)
train_docs.append(doc_tmp)
val_docs = []
for txt, con in val:
doc_tmp = Document(txt, con)
val_docs.append(doc_tmp)
test_docs = []
for txt, con in test:
doc_tmp = Document(txt, con)
test_docs.append(doc_tmp)
# file names
if not train_docs:
print('Error: Cannot train on 0 files. Terminating train.')
return 1
# Create a Machine Learning model
model = ClinerModel(use_lstm)
# Train the model using the Documents's data
model.train(train_docs, val=val_docs, test=test_docs)
# Pickle dump
print('\nserializing model to %s\n' % model_path)
with open(model_path, "wb") as m_file:
pickle.dump(model, m_file)