Skip to content

Latest commit

 

History

History
7 lines (6 loc) · 686 Bytes

README.md

File metadata and controls

7 lines (6 loc) · 686 Bytes

TensorFlowKerasIntro

Scripts to practice the basics of TensorFlow and Keras while building networks for image classification on the CIFAR (3 RGB channels) and MNIST (1 grayscale channel) image datasets.

  • Deep_CIFAR_KerasModel.py: CNN model (with grid search over hyper parameters) built with the Keras Model API applied on the CIFAR dataset.
  • Deep_CIFAR_LayersEstimators.py: CNN model built using tf.layers and the TF Estimator API applied on the CIFAR dataset.
  • Deep_CIFAR_LayersSession.py: CNN model built with tf.layers and tf.Session applied on the CIFAR dataset.
  • Deep_MNIST_TFcore.py: CNN model built with low-level TF API and tf.Session applied on the MNIST dataset.