autokeras

• Updated TF dependency version to 2.3.0.
• Bug fix for checkpoint not found during evaluate and predict.

Bug Fixes:

• Fixed fit the final model only one epoch after search all the trials.
• Fixed the checkpoint not found issue during fit.

New Features:

• Add pretrained XceptionNet and ResNet with ImageNet weights to the search space of image tasks.
• Enlarge the search space for more optimizers and laerning rates.
• Transformer model included in the search space.

API Changes:

• ResNetBlock and XceptionBlock pooling argument removed.
• All task APIs use overwrite=False by default.
• Change all dropout_rate arguments to dropout.

Bump dependency TensorFlow version to 2.2.0.
User now can use custom metrics and loss.
User now can specify the tuner to use for Task APIs like ImageClassifier.
Use Keras preprocessing layers for ImageAugmentation.
If not epochs is not specified, and validation_data is not provided,
it will use the best models best number of epochs for retraining the best model with the entire training set.
Bug fixes:
All bugs in the tutorials are fixed.
All the tutorials can run smoothly.
Breaking changes:
ImageAugmentation args updated.

Fixed the bug for low performance in the final model training.
Fixed the bug for does not support tf.data.Dataset for TextClassifier.
Improved performance for ImageClassifier and TextClassifier.

Issues:
It cannot save the preprocessing layers' weights.
The exported model has to be adapted manually if contains any preprocessing layer.
It should be fixed with TF 2.2.
We will have another release afterward.

Supporting Tensorflow Keras preprocessing layers.
The preprocessors are now exportable to Keras Model, too,
i.e., the entire model is exportable to Keras Model.
The exported model should have exactly the sample performances as the searched model.

Redesigned the API and system architecture based on KerasTuner 1.0 and TensorFlow 2.0.
Refer to the official website for more information. https://autokeras.com/

Use BERT to do the natural language tasks.
Pretrained models separated out to autokeras-pretrained.
Tabular module separated out to autokaggle.

Fixed some bugs.
Code refactored.

Added Predefined API.
Added pretrained models of face detection, object detection, voice generation.

Fixed the CUDA: out of memory bug.
Refined the interface for net_module.

Improved neural architecture search performance.

Fixed the bug in dependencies of setup.py

A comparatively stable version.

Fixed the memory limit bug.

Refined the output information on screen.

Updated output information.

Fixing the bugs caused by floating point.