Roadmap

Where we're headed

Phase 1: Precision & Performance

Float16 & BFloat16 Support

In Progress

Extend Candy with half-precision and brain floating-point support for memory-efficient training and inference.

  • float16 (IEEE 754 half-precision)
  • bfloat16 (Brain floating-point)
  • Mixed-precision training utilities
  • Automatic type casting
gocnn/candy
Phase 2: Inference & Interoperability

ONNX Bindings

Planned

ONNX ecosystem bindings for Go — load, run, and optimize ONNX models natively.

  • ONNX model parser and loader
  • ONNX Runtime Go bindings
  • Model optimization tools
  • Cross-platform inference
gocnn/goonx

TensorRT Bindings

Planned

NVIDIA TensorRT ecosystem bindings for high-performance inference on NVIDIA GPUs.

  • TensorRT core bindings
  • TensorRT-LLM for large language models
  • TensorRT-RTX for consumer GPUs
  • INT8/FP16 quantization support
gocnn/tensorrt
Phase 3: Reinforcement Learning

Sugar

Planned

Reinforcement learning framework for Go — environments, algorithms, and training utilities.

  • Gym-compatible environments
  • DQN, PPO, A2C algorithms
  • Multi-agent support
  • Distributed training
gocnn/sugar

Want to contribute?

We welcome contributions! Check out our GitHub for open issues and discussions.

Join the Discussion