Artelnics/opennn

Health v2 · maintenance only

Adoption not part of the score

1198 stars · 364 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

OpenNN is an open-source C++ library for building, training, and deploying neural networks for advanced analytics. It is dependency-free, optimized for execution speed and memory efficiency, and supports CPU (Eigen, MKL, oneDNN) and GPU (CUDA) backends with architectures including dense networks, LSTMs, CNNs, and transformers.

Use cases

• train neural networks for regression and classification in C++
• forecast time series with LSTM networks
• build CNN models for image and signal data
• run transformer models like BERT and GPT-2 with GPU acceleration
• customer segmentation and churn prediction for business intelligence
• early diagnosis and microarray analysis in health care
• predictive maintenance and performance optimization in engineering

When to choose

• you need a high-performance, dependency-free C++ neural network library
• you want CPU and GPU training/inference with CUDA, MKL, or oneDNN backends
• you are embedding machine learning into C++ engineering or analytics applications
• you need dense, recurrent, convolutional, or transformer architectures in one library

When to avoid

• you prefer Python-first ecosystems like PyTorch or TensorFlow
• you need a GUI-driven tool rather than a code library (consider Neural Designer)
• your project requires a permissive license, since OpenNN is LGPL-3.0
• you need a large community and extensive third-party ecosystem

Facets

library · maturity active

machine-learning deep-learning llm-inference image-processing math machine-learning deep-learning artificial-intelligence data-science computer-vision cpp cross-platform windows neural-networks c-plus-plus high-performance no-dependencies cpu-gpu transformers lstm cnn eigen cuda gpu linux macos

Member repositories

For agents

markdown · JSON · MCP: product_card(name="Artelnics/opennn")