sizif 0.0.2

Deep learning Keras models lifecycle management backup/restore nano framework

DL backup/restore nano framework

Makes it easy to start/stop/resume deep learning models training.

Current version supports only for Keras >= 2.2 models. You're welcome to contribute.

Usage

pip3 install sizif

Local filesystem Keras checkpoints backup:

from sizif.keras import KerasModelWrapper
from sizif.storage import FileCheckpointsMonitor

# your compiled Keras Model instance
model = build_model()  


# Snapshots monitor
# Different model architectures should have different version parameter
# other parameters similar to Keras ModelCheckpoint
cpm = FileCheckpointsMonitor(version=1,
                            file_template='weights.{epoch:03d}-vl{val_loss:.3f}.hdf5',
                            folder='./checkpoints',
                            rotate_number=5,
                            monitor='val_acc',
                            verbose=1,
                            save_best_only=False,
                            save_weights_only=True,
                            mode='auto',
                            period=1)

# Keras wrapper, proxies all calls to the model
# except fit and fit_generator — which are surrounded 
# by automated model state backup/recovery   
km = KerasModelWrapper(model, cpm)

# all method parameters ar proxied to Keras as is except callbacks
km.fit_generator(training_set_generator,
                 epochs=25,
                 validation_data=test_set_generator,
                 callbacks=[tboard])

See sources for detailed docstrings

TODO:

• FTP/S3/SFTP/Dropbox uploading monitors
• Tensorflow/Pytorch models support

Tests

python3 -m unittest 

Dependencies

• numpy ~> 1.15
• Keras ~> 2.2

License

This project is released under the MIT license.