/mnist

deep models for small image classification datasets

Primary LanguagePython

mnist

Supervised learning with feedforward deep models on small standard datasets. The main feature is the modular model import with model-specific arguments.

usage

python train.py -m cnn --m-activation sigmoid

TODO

  • add dataset CIFAR-10
  • add dataset SVHN
  • add ResNet-like model
  • add DenseNet-like model

author

Falcon Dai (me@falcondai.com)