https://github.com/mayer79/keras_examples

Simple Keras examples of deep learning
https://github.com/mayer79/keras_examples

Last synced: 15 days ago
JSON representation

Simple Keras examples of deep learning

• Host: GitHub
• URL: https://github.com/mayer79/keras_examples
• Owner: mayer79
• Created: 2018-06-04T07:07:45.000Z (over 6 years ago)
• Default Branch: master
• Last Pushed: 2018-07-24T13:24:15.000Z (over 6 years ago)
• Last Synced: 2024-10-04T12:57:06.233Z (3 months ago)
• Language: Python
• Size: 26.4 KB
• Stars: 1
• Watchers: 4
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# keras_examples

Simple Keras examples of deep learning inspired by the excellent book "Deep Learning with Python" from Francois Chollet, father of Keras.

All scripts are self-containing in the sense that it uses only data available through Keras. It will need a working installation of Keras package together with one of the supported back-ends (e.g. Tensorflow or Theano) and, for convolutional neural networks, a NVIDIA GPU (e.g. GTX 1080 Titan).

The scripts are all located in the "py" folder:

- binary_simple.py: Shows how to create a binary classifier

- multiclass_simple.py: Similar for a multiclass response

- regression_simple.py: Same for one numeric output aka regression

- mnist_simple.py: Simple (non-convolutional) net to classify hand-written digits 

- mnist_conv.py: Similar than above but now with convolutional layers

- mnist_conv_augmentation.py: Similar than above but now with image data augmentation through "flow"

- dog_detector.py: Similar than above but for a 120 class kaggle competition with dog breeds

- sentiment: Different RNNs for text input

- times_series: Multidimensional timeseries through LSTM