ScalaNet

Type-safe, high performance, distributed Neural networks in Scala (not Python, finally...).

Architecture Intro

Low level (linear algebra) operations powered by low level TensorFlow API (C, C++ bindings via JNI).

Scala used to build computation graphs and compile them into native tensor graphs. Compiled graphs are fully calculated in native code (on CPU, GPU or TPU) and only result is returned back via DirectBufferwhich points into native memory.

DirectBuffer is wrapped with Tensor object which allows to slice and read data in a convenient way (just like Breeze or Numpy does).

todo

Tasks

Tensor

DONE

Tensor Flow Low Level

1. Enhance Session
2. Add basics ops
3. Handle tensor graphs
4. Add gradient support to ops
5. Add the rest of the ops

CPU & GPU & TPU banchmarks

1. todo

Distributed Processing (Spark)

1. todo

Optimizers

1. todo

NN baseline + Multilayer Perceptron NN + Evaluation

1. todo

Additional Layers (dropout, etc.)

1. todo

Convolutional NN

1. todo

Recurent NN

1. todo