def main():
args = get_args()
experiment_name = 'yu4u'
early_stop_patience = config.EARLY_STOP_EPOCHS
PARAMS = {
'epoch_nr': args.nb_epochs,
'batch_size': args.batch_size,
'learning_rate': args.lr,
'early_stop': early_stop_patience,
'image_size': config.IMAGE_SIZE,
'network': args.model_name
}
neptune.init(project_qualified_name='4ND4/sandbox')
neptune_tb.integrate_with_keras()
result = neptune.create_experiment(name=experiment_name, params=PARAMS)
name = result.id
print(name)
image_path = args.pred_dir
model_name = args.model_name
batch_size = args.batch_size
nb_epochs = args.nb_epochs
lr = args.lr
image_size = config.IMAGE_SIZE
train_path = os.path.join(image_path, 'train')
validation_path = os.path.join(image_path, 'validation')
test_path = os.path.join(image_path, 'test')
train_gen, val_gen, test_gen = getdata(train_path, validation_path,
test_path)
model = get_model(model_name=model_name,
image_size=image_size,
number_classes=nb_classes)
sgd = SGD(lr=0.1, momentum=0.9, nesterov=True)
model.compile(optimizer=sgd,
loss="categorical_crossentropy",
metrics=[age_mae])
model.summary()
output_dir = Path(__file__).resolve().parent.joinpath(args.output_dir)
if not output_dir.exists():
output_dir.mkdir(parents=True)
if not os.path.exists('checkpoints/{}'.format(name)):
os.mkdir('checkpoints/{}'.format(name))
callbacks = [
EarlyStopping(monitor='val_age_mae',
mode='min',
verbose=1,
patience=early_stop_patience),
LearningRateScheduler(schedule=Schedule(nb_epochs, initial_lr=lr)),
ModelCheckpoint(os.path.join(output_dir, name) +
"/weights.{epoch:03d}-{val_loss:.3f}-{"
"val_age_mae:.3f}.hdf5",
monitor="val_age_mae",
verbose=1,
save_best_only=True,
mode="min")
]
hist = model.fit_generator(train_gen,
steps_per_epoch=train_gen.samples // batch_size,
validation_data=val_gen,
epochs=nb_epochs,
verbose=1,
callbacks=callbacks)
np.savez(str(output_dir.joinpath("history_{}.npz".format(name))),
history=hist.history)
def main():
nb_epochs = config.MAXIMUM_EPOCHS
# changing params with winning hyper-params
# lr = 0.1
# momentum = 0.9
# batch_size = config.batch_size
image_size = config.IMAGE_SIZE
batch_size = 96
lr = 0.0008828491766861475
momentum = 0.878361507747756
image_size = 224
model_name = 'ResNet50'
output_dir = 'checkpoints'
experiment_name = 'DeepUAge2.0'
early_stop_patience = config.EARLY_STOP_EPOCHS
train_path = os.path.join(image_directory, 'train')
validation_path = os.path.join(image_directory, 'validation')
# test_path = os.path.join(image_directory, 'test')
PARAMS = {
'epoch_nr': nb_epochs,
'batch_size': batch_size,
'learning_rate': lr,
'momentum': momentum,
'early_stop': early_stop_patience,
'image_size': image_size,
'network': model_name,
'monitor': monitor,
'dataset_path': image_directory
}
if log_results:
neptune.init(project_qualified_name='4ND4/sandbox')
neptune_tb.integrate_with_keras()
result = neptune.create_experiment(name=experiment_name, params=PARAMS)
name = result.id
else:
name = 'debug'
print(name)
train_gen, val_gen = getdata(train_path, validation_path)
model = get_model(model_name=model_name,
image_size=image_size,
number_classes=nb_classes)
sgd = SGD(lr=lr, momentum=momentum, nesterov=True)
model.compile(optimizer=sgd,
loss="categorical_crossentropy",
metrics=[age_mae])
model.summary()
output_dir = Path(__file__).resolve().parent.joinpath(output_dir)
if not output_dir.exists():
output_dir.mkdir(parents=True)
if not os.path.exists('checkpoints/{}'.format(name)):
os.mkdir('checkpoints/{}'.format(name))
callbacks = [
EarlyStopping(monitor=monitor,
mode='min',
verbose=1,
patience=early_stop_patience),
ModelCheckpoint(os.path.join(output_dir, name) +
"/weights.{epoch:03d}-{val_loss:.3f}-{"
"val_age_mae:.3f}.hdf5",
monitor=monitor,
verbose=1,
save_best_only=True,
mode="min")
]
hist = model.fit_generator(generator=train_gen,
epochs=nb_epochs,
validation_data=val_gen,
verbose=1,
callbacks=callbacks)
np.savez(str(output_dir.joinpath("history_{}.npz".format(name))),
history=hist.history)
PARAMS = {
'epoch_nr': EPOCHS,
'batch_size': 64,
'learning_rate': 0.006,
'input_shape': (512, 32, 3),
'early_stop': 20
}
# start experiment
name = 'resnet50-experiment'
if LOG_NEPTUNE:
neptune.init(project_qualified_name='4ND4/sandbox')
neptune_tb.integrate_with_keras()
result = neptune.create_experiment(name=name, params=PARAMS)
name = result.id
def getdata():
_data = np.load('output/DeepUAge-faces-embeddings_{}_{}.npz'.format(
VECTOR_SIZE, FACE_DETECTION))
embeddings_val_npz_path = 'output/DeepUAge-val-faces-embeddings-{}-{}.npz'.format(
VECTOR_SIZE, FACE_DETECTION)
trainX, trainy, valX, valy = _data['arr_0'], _data['arr_1'], _data[
'arr_2'], _data['arr_3']
# print('Dataset: train=%d, test=%d' % (trainX.shape[0], valX.shape[0]))
def main():
nb_epochs = config.MAXIMUM_EPOCHS
batch_size = config.batch_size
lr = 0.1
momentum = 0.9
model_name = 'ResNet50'
image_size = config.IMAGE_SIZE
output_dir = 'checkpoints'
experiment_name = 'yu4u'
early_stop_patience = config.EARLY_STOP_EPOCHS
PARAMS = {
'epoch_nr': nb_epochs,
'batch_size': batch_size,
'learning_rate': lr,
# 'input_shape': (512, 32, 3),
'early_stop': early_stop_patience,
'image_size': config.IMAGE_SIZE,
'network': model_name
}
if config.LOG_RESULTS:
neptune.init(project_qualified_name='4ND4/sandbox')
neptune_tb.integrate_with_keras()
result = neptune.create_experiment(name=experiment_name, params=PARAMS)
name = result.id
else:
name = 'no_log'
train_gen = FaceGenerator(image_directory,
batch_size=batch_size,
image_size=image_size,
number_classes=nb_classes)
val_gen = ValGenerator(image_directory,
batch_size=batch_size,
image_size=image_size,
number_classes=nb_classes)
model = get_model(model_name=model_name,
image_size=image_size,
number_classes=nb_classes)
sgd = SGD(lr=lr, momentum=momentum, nesterov=True)
model.compile(optimizer=sgd,
loss="categorical_crossentropy",
metrics=[age_mae])
model.summary()
output_dir = Path(__file__).resolve().parent.joinpath(output_dir)
if not output_dir.exists():
output_dir.mkdir(parents=True)
if not os.path.exists('checkpoints/{}'.format(name)):
os.mkdir('checkpoints/{}'.format(name))
callbacks = [
EarlyStopping(monitor='val_age_mae',
mode='min',
verbose=1,
patience=early_stop_patience),
LearningRateScheduler(schedule=Schedule(nb_epochs, initial_lr=lr)),
ModelCheckpoint(os.path.join(output_dir, name) +
"/weights.{epoch:03d}-{val_loss:.3f}-{"
"val_age_mae:.3f}.hdf5",
monitor="val_age_mae",
verbose=1,
save_best_only=True,
mode="min")
]
hist = model.fit_generator(generator=train_gen,
epochs=nb_epochs,
validation_data=val_gen,
verbose=1,
callbacks=callbacks)
np.savez(str(output_dir.joinpath("history_{}.npz".format(name))),
history=hist.history)