Tenncor

Synopsis

Tenncor libraries help developers build and evaluate tensor equations and its derivatives. A tensor is an N-dimensional container that organizes its content by some shape. An M by N matrix for instance, is a 2-dimensional tensor with a shape of [N, M] (according to Tenncor's x-y-z-... coordinate notation).

High-level diagram available: https://drive.google.com/file/d/1PrsFa7Duj4Whlu_m0lmFr5JGikGnU3gC/view?usp=sharing

Components

• TEQ (Tensor EQuations)

This module supplies syntax tree for equation and generates derivative. Constraints to the equation is limited to each tensor's shape.

• DBG (Debug)

This module is contains debug libraries for TEQ Graphs.

• ETEQ (Eigen TEQ)

This module is implements basic operations for Tenncor's TEQ Tensor objects generated through pybinder. Additionally, ETEQ also defines data format and (de)serialization methods required by PBM.

• OPT (OPTimizer)

This module specifies graph optimization through TEQ's visitor pattern.

• PBM (ProtoBuf Marshaller)

This module marshals any TEQ graph, but requires data serialization functors when saving and loading.

• TAG (TAGger)

This module tags TEQ tensors with labels.

• CCUR (ConCURrent session)

This module implements session that updates graph nodes concurrently

• LAYR (LAYeR models)

This module implements common machine learning models

Tools and utility

• DBG (DeBuGger)

Building

Tenncor uses bazel 0.28+.

Download bazel: https://docs.bazel.build/versions/master/install.html