/nn_basic

Some NN implementations for image classification and neural machine translation

Primary LanguageJupyter NotebookMIT LicenseMIT

nn_basic

This repository contains some basic NN implementations for image classification and neural machine translation.

1. Fully-connected neural network from scratch

Codes

  • demo_fc_from_scratch.ipynb

Requirements

  • Python 3
  • NumPy

2. Image classification with a basic CNN

  • demo_cnn_basic_cifar10.ipynb

Requirements

  • Python 2.7
  • PyTorch
  • NumPy

3. Image classification with ResNet

  • demo_cnn_resnet_cifar10.ipynb

Requirements

  • Python 2.7
  • PyTorch
  • NumPy

4. Neural machine translation with seq2seq model

  • demo_nmt_seq2seq.ipynb

Requirements

  • Python 2.7
  • PyTorch
  • NumPy
  • NLTK

5. CNN to classify large 3D regions

  • demo_cnn_3d_classification.ipynb

Requirements

  • Python 2.7
  • Keras (TensorFlow backend)
  • nibabel (interface to nifti files)
  • scikit-learn