ChainGraph is a source available, type-safe flow-based programming framework for building AI LLM agents with complex logic and dynamic computational graphs.

Computational graph-based discrete choice models

Python library providing a collection of functions realizing common computer vision functionality, based on OpenCV and NumPy.

A general purpose framework for building and running computational graphs.

Network-wide estimation of traffic flow and travel time with data-driven macroscopic models

A comprehensive C++ framework for automatic differentiation supporting both forward and reverse modes with exceptional performance. Designed with solid software engineering principles, AutoDiff provides exact derivatives of arbitrary functions with machine precision through dual numbers and computational graphs.

A computational graph for time-series processing.

🎢 IaaS visual editor to create & deploy data processing pipelines - python, rmq, react, meteorjs

This is an experiment version of calibrating origin-destination matrix estimation using link traffic counts

Code for "Can We Scale Transformers to Predict Parameters of Diverse ImageNet Models?" [ICML 2023]

Yet another tensor automatic differentiation framework

(Fixed) Computational Graph Library in Rust

RNN in Julia for MNIST digit recognition implemented with automatic differentiation. Over 96% accuracy.

See how backpropagation and chain rule work in neural networks

Computation Graph framework implemented using only NumPy

Parameter Estimation of LOGIT-based Stochastic User Equilibrium models using computational graphs and day-to-day system-level data

a compact tensor library capable of training deep neural networks on both cpu and cuda devices

A deep learning library for golang

A graph-oriented algorithmic engine

TensorFlow's very distant and not so bright cousin