OpenNN

OpenNN is an open-source neural networks library for C++ designed for machine learning tasks such as regression, classification, forecasting, and association. It claims higher capacity (loading data sets 1.8 times larger than TensorFlow and PyTorch) and faster training (2.5 times faster than PyTorch and 1.55 times faster than TensorFlow). Developed by Artelnics, a consulting company specializing in AI and big data, OpenNN is used by enterprise customers including Philips, Seat, Airbus, and Acciona. The library provides sophisticated algorithms and utilities for building neural network models without the need for a graphical interface, making it suitable for developers and data scientists who prefer programmatic control. Additionally, Artelnics offers consulting services and a related no-code tool, Neural Designer, built on OpenNN.

Key Benefits

• Open-source neural networks library
• Higher capacity: loads data sets 1.8 times larger than TensorFlow and PyTorch
• Faster training: 2.5 times faster than PyTorch and 1.55 times faster than TensorFlow
• Supports regression, classification, forecasting, and association tasks