tranlethaison / NumpyNeuralNet

Neural Networks and training algorithms in Numpy, for learning purpose.

Repository from Github https://github.comtranlethaison/NumpyNeuralNetRepository from Github https://github.comtranlethaison/NumpyNeuralNet

NumpyNeuralNet

As the man himself said "What I cannot create, I do not understand.".
So I try to built Neural Networks and training algorithms using only Numpy.

Densely connected NN

Train a MNIST model with [output_activation]_[Loss].

  • Sigmoid - MSE

    $ python test/dense.py sigmoid_mse train
  • Sigmoid - Cross-entropy

    $ python test/dense.py sigmoid_crossentropy train
  • Softmax - Log-likelihood

    $ python test/dense.py softmax_loglikelihood train

About

Neural Networks and training algorithms in Numpy, for learning purpose.

License:MIT License


Languages

Language:Python 100.0%