def updateStatus(serviceid, train_ner, train_ir):
manager = ProjectManager()
query = {"serviceid": serviceid}
config = manager.find_model(query)
if config is not None:
if train_ner is True:
document = {
"$set": {
"ner.status":
ProjectManager.STATUS_HOLD,
"ner.status_message":
"Awaiting the completion of entity training."
}
}
manager.update_config(query, document)
if train_ir is True:
document = {
"$set": {
"ir.status":
ProjectManager.STATUS_HOLD,
"ir.status_message":
"Awaiting the completion of intent training."
}
}
manager.update_config(query, document)
def update_last_access_to_predict_api(serviceid):
query = {"serviceid": serviceid}
document = {
"$set": {
"lastAccessed": datetime.datetime.utcnow(),
}
}
manager = ProjectManager()
manager.update_config(query, document)
def on_post(self, req, resp):
doc = req.context['doc'] or {}
try:
home_dir = os.path.expanduser('~')
model_store_path = os.path.join(home_dir + os.sep +
'.verbis/store')
model_store = get_model_store()
manager = ProjectManager()
serviceid = doc['serviceid']
query = {'serviceid': serviceid}
data = manager.find_model(query)
engine = get(data, "engine", "ICE")
model_name = get_model_name(serviceid, MODEL_TYPE_NER, engine)
model_lists = model_store.get_active_models()
logger.info("model lists - %s" % model_lists)
if model_name in model_lists:
model_store.unpublish(serviceid, MODEL_TYPE_NER, engine)
model_name = get_model_name(serviceid, MODEL_TYPE_IR, engine=None)
logger.info(model_name)
if model_name in model_lists:
model_store.unpublish(serviceid, MODEL_TYPE_IR, engine=None)
logger.info("path is %s" %
os.path.join(model_store_path + os.sep + serviceid))
# delete model file from local
if os.path.exists(
os.path.join(model_store_path + os.sep + serviceid)):
# print "path exists, so do rm."
shutil.rmtree(
os.path.join(model_store_path + os.sep + serviceid))
# delete model file from minio
VerbisStore().remove_models_from_remote(serviceid)
logger.info("files removed successfully")
# update DB
document = {"$set": {"ner.status": 'new', "ir.status": 'new'}}
manager.update_config(query, document)
except AssertionError as ae:
logger.exception(ae, exc_info=True)
logger.error(traceback.format_exc())
raise falcon.HTTPPreconditionFailed(
'Service publish condition failed', traceback.format_exc())
except Exception as ex:
logger.exception(ex, exc_info=True)
logger.error(traceback.format_exc())
description = 'Internal Server Error, Please try again later'
raise falcon.HTTPServiceUnavailable('Service Outage', description,
30)
resp.set_header('X-Powered-By', 'USTGlobal Verbis')
resp.status = falcon.HTTP_200
def put_corenlp_modelname(serviceid, model_name):
manager = ProjectManager()
query = {"serviceid": serviceid}
document = {"$set": {"corenlp_model_name": model_name.split("/")[-1]}}
manager.update_config(query, document)
def deploy(self):
"""
:param doc:
:param n_test_percent:
:return:
"""
manager = ProjectManager()
query = {
"serviceid": self.serviceid
}
config = manager.find_model(query)
if config is not None:
try:
trained_data, df = get_ir_dataset(self.serviceid, self.logger)
self.logger.info("Unique labels %s" % np.unique(df.intent.tolist()))
group = df.groupby(['intent']).agg('count')
stats = group.reset_index().to_json(orient="records")
useSelector = False
if (len(group) > 1):
useSelector = True
self.logger.info(stats)
document = {
"$set": {
"ir.status": ProjectManager.STATUS_TRAINING,
"ir.status_message": "Intent training is in progress.",
"ir.dataset.stats": stats,
"ir.last_trained": datetime.datetime.utcnow()
}
}
manager.update_config(query, document)
ir = IntentRecognizer(DEFAULT_CONFIG, serviceid=self.serviceid, useSelector=useSelector)
self.logger.info("Starting fitting for deployment")
ir.fit(df, df.intent)
self.logger.info("Fitting for deployment completed")
VerbisStore().save_ir(ir)
###############MINIOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO##################
# send file to minio server
# no engine. .dat - extension
VerbisStore().save_ir_minio(ir)
document = {
"$set": {
"utterances": trained_data,
}
}
data_manager = DatasourceManager()
data_manager.update_datasource(query, document)
document = {
"$set": {
"ir.status": ProjectManager.STATUS_TRAINED,
"ir.status_message": "The Intent model has been successfully trained",
"ir.logs.deploy": ""
}
}
manager.update_config(query, document)
except (RuntimeError, ValueError, Exception) as e:
self.logger.error(e)
message = e
if (e == "After pruning, no terms remain. Try a lower min_df or a higher max_df."
or e == "max_df corresponds to < documents than min_df"):
message = "Sufficient vocabulary to build the model is not available. Please add more utterances."
elif e == "Invalid type float for labels":
message = "Add more intents for intent training"
document = {
"$set": {
"ir.status": ProjectManager.STATUS_TRAINING_FAILED,
"ir.status_message": message,
"ir.logs.deploy": self.logger.handlers[-1].logs
}
}
manager.update_config(query, document)
traceback.print_exc()
else:
description = 'Unable to find project_config with given id.' \
'Please check your request params and retry'
self.logger.error(description)
def train(self, train_intent):
"""
:param doc:
:param n_test_percent:
:return:
"""
manager = ProjectManager()
query = {"serviceid": self.serviceid}
config = manager.find_model(query)
if config is not None:
try:
document = {
"$set": {
"ner.status": ProjectManager.STATUS_TRAINING,
"ner.status_message":
"Entity training is in progress.",
"ner.last_trained": datetime.datetime.utcnow()
}
}
if (train_intent is True):
document = {
"$set": {
"ner.status": ProjectManager.STATUS_TRAINING,
"ner.status_message":
"Entity training is in progress.",
"ir.status": ProjectManager.STATUS_HOLD,
"ir.status_message":
"Awaiting the completion of entity training.",
"ner.last_trained": datetime.datetime.utcnow()
}
}
manager.update_config(query, document)
# starting actual training
data_manager = DatasourceManager()
self.logger.info("Starting training of service %s" %
self.serviceid)
corpus = data_manager.find_model(query)
custom_entity_model = get(config, "custom_entity_model")
entity_recognizer = self.instantiate_trainer(
custom_entity_model)
trained_utterances = entity_recognizer.train(corpus)
if entity_recognizer.get_engine(
) not in get_all_corenlp_engines():
VerbisStore().save_ner(entity_recognizer,
model_type=MODEL_TYPE_NER)
###############MINIOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO##################
# send file to minio server
VerbisStore().save_ner_minio(entity_recognizer,
model_type=MODEL_TYPE_NER)
document = {
"$set": {
"utterances": trained_utterances,
}
}
data_manager.update_datasource(query, document)
document = {
"$set": {
"ner.status": ProjectManager.STATUS_TRAINED,
"ner.status_message":
"Entity training completed successfully.",
"ner.logs.train": ""
}
}
manager.update_config(query, document)
self.logger.info(
"Completed training entity recognizer for service %s" %
self.serviceid)
except (RuntimeError, Exception) as ex:
self.logger.exception(ex, exc_info=True)
self.logger.error(traceback.format_exc())
if ex == "Cannot have number of folds n_folds=3 greater than the number of samples: 2.":
ex = "Add more utterances for entity training"
document = {
"$set": {
"ner.status": ProjectManager.STATUS_TRAINING_FAILED,
"ner.status_message": ex,
"ner.logs.train": self.logger.handlers[-1].logs
}
}
manager.update_config(query, document)
else:
description = 'Unable to find project_config with given id.' \
'Please check your request params and retry'
self.logger.error(description)