def tt_trial(args, reporter):
""" Training and evaluation function used during a single trial of HPO """
try:
model, args, util_args = model_trial.prepare_inputs(args=args)
fit_model_args = dict(X=util_args.X,
y=util_args.y,
X_val=util_args.X_val,
y_val=util_args.y_val,
**util_args.get('fit_kwargs', dict()))
predict_proba_args = dict(X=util_args.X_val)
model = model_trial.fit_and_save_model(
model=model,
params=args,
fit_args=fit_model_args,
predict_proba_args=predict_proba_args,
y_val=util_args.y_val,
time_start=util_args.time_start,
time_limit=util_args.get('time_limit', None),
reporter=reporter)
except Exception as e:
if not isinstance(e, TimeLimitExceeded):
logger.exception(e, exc_info=True)
reporter.terminate()
else:
reporter(epoch=model.params['epochs'] + 1,
validation_performance=model.val_score)
pass
def tabular_pytorch_trial(args, reporter):
""" Training and evaluation function used during a single trial of HPO """
try:
model, args, util_args = model_trial.prepare_inputs(args=args)
train_dataset = TabularPyTorchDataset.load(util_args.train_path)
val_dataset = TabularPyTorchDataset.load(util_args.val_path)
y_val = val_dataset.get_labels()
fit_model_args = dict(X=train_dataset,
y=None,
X_val=val_dataset,
**util_args.get('fit_kwargs', dict()))
predict_proba_args = dict(X=val_dataset)
model_trial.fit_and_save_model(model=model,
params=args,
fit_args=fit_model_args,
predict_proba_args=predict_proba_args,
y_val=y_val,
time_start=util_args.time_start,
time_limit=util_args.get(
'time_limit', None),
reporter=reporter)
except Exception as e:
if not isinstance(e, TimeLimitExceeded):
logger.exception(e, exc_info=True)
reporter.terminate()
def lgb_trial(args, reporter):
""" Training script for hyperparameter evaluation of Gradient Boosting model """
try:
model, args, util_args = model_trial.prepare_inputs(args=args)
try_import_lightgbm()
import lightgbm as lgb
dataset_train = lgb.Dataset(util_args.directory + util_args.dataset_train_filename)
dataset_val = lgb.Dataset(util_args.directory + util_args.dataset_val_filename)
X_val, y_val = load_pkl.load(util_args.directory + util_args.dataset_val_pkl_filename)
fit_model_args = dict(dataset_train=dataset_train, dataset_val=dataset_val, **util_args.get('fit_kwargs', dict()))
predict_proba_args = dict(X=X_val)
model_trial.fit_and_save_model(model=model, params=args, fit_args=fit_model_args, predict_proba_args=predict_proba_args, y_val=y_val,
time_start=util_args.time_start, time_limit=util_args.get('time_limit', None), reporter=reporter)
except Exception as e:
if not isinstance(e, TimeLimitExceeded):
logger.exception(e, exc_info=True)
reporter.terminate()