ConvNetSwift

Swift port of ConvnetJS - a library for training Deep Learning models (mainly Neural Networks).

Updated to Swift 3.

Available NN layer types:

• [x] Fully Connected
• [x] Convolutional
• [x] Dropout
• [x] ReLU
• [x] Sigmoid
• [x] Maxout
• [x] Hyperbolic Tangent
• [x] Local Response Normalization
• [x] Max Pooling
• [x] Output: Softmax, Regression, SVM

Optimization methods ('trainers'/'solvers'):

• [x] Stochastic Gradient Descent
• [x] Adam
• [x] Adaptive Gradient
• [x] WindowGrad
• [x] AdaDelta
• [x] Nesterov’s Accelerated Gradient

Reinforcement learning:

• [x] Q-learning

Acceleration:

• [ ] Accelerate framework
• [ ] Metal

This is not a production-ready library. It is a personal research project and work in progress.