def __init__(self, *args, **kws):
KerasModelCheckpoint.__init__(self, *args, **kws)
if self.filepath.startswith("gs://"):
self.on_epoch_end = self._gcp_on_epoch_end
self._original_filepath = self.filepath
self._temp_file = tempfile.NamedTemporaryFile()
self.filepath = self._temp_file.name
def __init__(self,
directory,
filename,
monitor='val_loss',
verbose=0,
save_best_only=False,
save_weights_only=False,
mode='auto',
period=1):
# make folder with the current time as name
now = datetime.datetime.now()
current_time = "{}_{}_{}_{}_{}_{}".format(now.day, now.month, now.year,
now.hour, now.minute,
now.second)
constants.SAVE_DIR = os.path.join(directory, current_time)
create_folder(constants.SAVE_DIR)
ModelCheckpoint.__init__(self,
os.path.join(constants.SAVE_DIR, filename),
monitor=monitor,
save_best_only=save_best_only,
save_weights_only=save_weights_only,
mode=mode,
period=period)
def __init__(self,
filepath,
monitor='val_loss',
verbose=0,
save_best_only=False,
mode='auto',
start_epoch=0):
ModelCheckpoint.__init__(self,
filepath,
monitor=monitor,
verbose=verbose,
save_best_only=save_best_only,
mode=mode)
self.start_epoch = start_epoch
def __init__(self, filepath, save_best_only=True, training_set=(None, None), testing_set=(None, None), folder=None, cost_string="log_loss", save_training_dataset=False, verbose=1):
ModelCheckpoint.__init__(self, filepath=filepath, save_best_only=save_best_only, verbose=1)
self.training_x, self.training_y = training_set
self.testing_x, self.testing_id, = testing_set
self.folder = folder
self.save_training_dataset = save_training_dataset
if cost_string == "log_loss":
self.cost_function = cost_string
elif cost_string == "auc":
self.cost_function = roc_auc_score
else:
log("Found undefined cost function - {}".format(cost_string), ERROR)
raise NotImplementError
def __init__(self,
name,
directory='',
associated_trial=None,
monitor='val_loss',
verbose=0,
save_best_only=True,
mode='auto'):
self.name = name
if (associated_trial != None):
self.smartDir = associated_trial.get_path()
self.checkpointFilename = self.smartDir + "weights.h5"
self.historyFilename = self.smartDir + "history.json"
else:
self.smartDir = directory + 'SmartCheckpoint/'
self.checkpointFilename = self.smartDir + name + "_weights.h5"
self.historyFilename = self.smartDir + name + "_history.json"
self.startTime = 0
# self.max_epoch = max_epoch
self.histobj = History()
histDict = {}
try:
histDict = json.load(open(self.historyFilename, "rb"))
print('Sucessfully loaded history at ' + self.historyFilename)
except (IOError, EOFError):
print('Failed to load history at ' + self.historyFilename)
self.histobj.history = histDict
self.elapse_time = histDict.get("elapse_time", 0)
ModelCheckpoint.__init__(self, self.checkpointFilename, monitor,
verbose, save_best_only, mode)
metric_history = histDict.get(monitor, None)
if (metric_history != None):
best = metric_history[0]
for metric in metric_history:
if self.monitor_op(metric, self.best):
self.best = metric
def __init__(self,
filepath,
base_model,
monitor='val_loss',
verbose=0,
save_best_only=False,
save_weights_only=False,
mode='auto',
period=1):
ModelCheckpoint.__init__(self,
filepath,
monitor=monitor,
verbose=verbose,
save_best_only=save_best_only,
save_weights_only=save_weights_only,
mode=mode,
period=period)
self.base_model = base_model
def __init__(self, *args, **kwargs):
self.model_name_ = kwargs.pop('name', 'unknown')
self.save_every_k_epochs_ = kwargs.pop('save_every_k_epochs', 100)
ModelCheckpoint.__init__(self, *args, **kwargs)
