gorgonia/gorgonia

Health v2 · maintenance only

Adoption not part of the score

5929 stars · 450 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

Gorgonia is a Go library for machine learning that lets you define and evaluate mathematical equations over multidimensional arrays using a computation graph, similar in spirit to Theano and TensorFlow. It supports automatic and symbolic differentiation, gradient descent optimization, neural network construction, and CUDA/GPGPU computation.

Use cases

• build neural networks in go
• automatic differentiation library for go
• train machine learning models without python
• compute gradients of mathematical expressions
• gpu-accelerated tensor math in go
• deploy deep learning models in go services
• experiment with non-standard learning algorithms

When to choose

• your stack is Go and you want production ML without rewriting models in another language
• you need a low-level, Theano-style computation graph library in Go
• you want simple compilation and deployment of ML systems
• you're exploring experimental learning algorithms like evolutionary or neo-hebbian learning

When to avoid

• you need the mature ecosystem, tooling, and community of PyTorch or TensorFlow
• you require OpenCL support or distributed training today
• you want cutting-edge model architectures and pretrained models
• you need rapid experimentation typical of Python-based data science workflows

Facets

library · maturity maintenance

machine-learning deep-learning math gpu-computing machine-learning deep-learning artificial-intelligence go windows computation-graph automatic-differentiation neural-networks tensors gradient-descent theano-like algorithms linux macos gpu

Member repositories

For agents

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