def init():
global model
global batch_size
# Set up logging
_set_logging_parameters(TASK_TYPE, {})
parser = argparse.ArgumentParser(
description="Retrieve model_name and batch_size from arguments.")
parser.add_argument('--model_name', dest="model_name", required=True)
parser.add_argument('--batch_size',
dest="batch_size",
type=int,
required=False)
args, _ = parser.parse_known_args()
batch_size = args.batch_size
model_path = os.path.join(Model.get_model_path(args.model_name),
'model.pt')
print(model_path)
try:
logger.info("Loading model from path: {}.".format(model_path))
model_settings = {}
model = load_model(TASK_TYPE, model_path, **model_settings)
logger.info("Loading successful.")
except Exception as e:
logging_utilities.log_traceback(e, logger)
raise
def init():
global model
# This name is model.id of model that we want to deploy deserialize the model file back
# into a sklearn model
model_path = os.path.join(os.getenv('AZUREML_MODEL_DIR'), 'model.pkl')
try:
model = joblib.load(model_path)
except Exception as e:
path = os.path.normpath(model_path)
path_split = path.split(os.sep)
log_server.update_custom_dimensions({'model_name': path_split[1], 'model_version': path_split[2]})
logging_utilities.log_traceback(e, logger)
raise
def init():
global model
#model_path = Model.get_model_path("best_hyperdrive_model")
model_path = os.path.join(os.getenv('AZUREML_MODEL_DIR'), 'model.joblib')
path = os.path.normpath(model_path)
path_split = path.split(os.sep)
log_server.update_custom_dimensions({
'model_name': path_split[1],
'model_version': path_split[2]
})
try:
logger.info("Loading model from path.")
model = joblib.load(model_path)
logger.info("Loading successful.")
except Exception as e:
logging_utilities.log_traceback(e, logger)
raise
def init():
global model
# This name is model.id of model that we want to deploy deserialize the model file back
# into a sklearn model
model_base_path = os.path.join(os.getenv('AZUREML_MODEL_DIR'), 'outputs')
model_path = os.path.join(model_base_path, 'model.pkl')
path = os.path.normpath(model_path)
path_split = path.split(os.sep)
log_server.update_custom_dimensions({
'model_name': path_split[-3],
'model_version': path_split[-2]
})
try:
logger.info("Loading model from path.")
model = joblib.load(model_path)
logger.info("Loading successful.")
except Exception as e:
logging_utilities.log_traceback(e, logger)
raise
def init():
global model
# This name is model.id of model that we want to deploy deserialize the model file back
# into a sklearn model
#model_path = os.path.join(os.getenv('AZUREML_MODEL_DIR'), 'automl_bestmodel.pkl')
#model_path= Model.get_model_path(model_name="automl_bestmodel.pkl")
#logger.error(model_path+" Muzammil1")
model_path = os.path.join(os.getenv('AZUREML_MODEL_DIR'), 'model.pkl')
#aar= os.listdir(model_path)
#print(aar)
#logger.error(aar+" Muzammil2")
path = os.path.normpath(model_path)
path_split = path.split(os.sep)
log_server.update_custom_dimensions({'model_name': path_split[1], 'model_version': path_split[2]})
try:
logger.info("Loading model from path.")
model = joblib.load(model_path)
logger.info("Loading successful.")
except Exception as e:
logging_utilities.log_traceback(e, logger)
raise
def run(input_data):
entry_script = EntryScript()
logger = entry_script.logger
os.makedirs('./outputs', exist_ok=True)
resultList = []
model_name = None
current_run = None
error_message = None
error_code = None
error_type = None
tags_dict = None
for file in input_data:
logs = []
date1 = datetime.datetime.now()
logger.info('start (' + file + ') ' + str(datetime.datetime.now()))
file_path = file
file_name, file_extension = os.path.splitext(
os.path.basename(file_path))
try:
if file_extension.lower() == ".parquet":
data = pd.read_parquet(file_path)
else:
data = pd.read_csv(file_path, parse_dates=[timestamp_column])
tags_dict = {'ModelType': 'AutoML'}
for column_name in group_column_names:
tags_dict.update({
column_name:
str(data.iat[0, data.columns.get_loc(column_name)])
})
if args.retrain_failed_models:
logger.info('querying for existing models')
try:
tags = [[k, v] for k, v in tags_dict.items()]
models = Model.list(current_step_run.experiment.workspace,
tags=tags,
latest=True)
if models:
logger.info("model already exists for the dataset " +
models[0].name)
logs = compose_logs(file_name, models[0], date1)
resultList.append(logs)
continue
except Exception as error:
logger.info('Failed to list the models. ' +
'Error message: ' + str(error))
tags_dict.update({'InputData': file_path})
tags_dict.update({'StepRunId': current_step_run.id})
tags_dict.update({'RunId': current_step_run.parent.id})
# train model
fitted_model, current_run = train_model(file_path, data, logger)
model_string = '_'.join(
str(v) for k, v in sorted(tags_dict.items())
if k in group_column_names).lower()
logger.info("model string to encode " + model_string)
sha = hashlib.sha256()
sha.update(model_string.encode())
model_name = 'automl_' + sha.hexdigest()
tags_dict.update({'Hash': sha.hexdigest()})
try:
logger.info('done training')
print('Trained best model ' + model_name)
logger.info(fitted_model)
logger.info(model_name)
logger.info('register model')
current_run.register_model(model_name=model_name,
description='AutoML',
tags=tags_dict)
print('Registered ' + model_name)
except Exception as error:
error_type = ErrorTypes.Unclassified
error_message = 'Failed to register the model. ' + 'Error message: ' + str(
error)
from azureml.automl.core.shared import logging_utilities
logging_utilities.log_traceback(error, None)
logger.info(error_message)
date2 = datetime.datetime.now()
logs.append('AutoML')
logs.append(file_name)
logs.append(current_run.id)
logs.append(current_run.get_status())
logs.append(model_name)
logs.append(tags_dict)
logs.append(str(date1))
logs.append(str(date2))
logs.append(error_type)
logs.append(error_code)
logs.append(error_message)
logger.info('ending (' + file_path + ') ' + str(date2))
# 10.1 Log the error message if an exception occurs
except (ClientException, AutoMLException) as error:
date2 = datetime.datetime.now()
error_message = 'Failed to train the model. ' + 'Error : ' + str(
error)
logs.append('AutoML')
logs.append(file_name)
if current_run:
logs.append(current_run.id)
logs.append(current_run.get_status())
else:
logs.append(current_run)
logs.append('Failed')
logs.append(model_name)
logs.append(tags_dict)
logs.append(str(date1))
logs.append(str(date2))
if isinstance(error, AutoMLException):
logs.append(error.error_type)
else:
logs.append(None)
logs.append(get_error_code(error))
logs.append(error_message)
logger.info(error_message)
logger.info('ending (' + file_path + ') ' + str(date2))
resultList.append(logs)
result = pd.DataFrame(data=resultList)
return result
