R
RUMUS
Get Started
Docs
Getting StartedTensorsAutogradNeural NetworksOptimizersGPU Acceleration
Community

Introduction

  • Getting Started

Core Concepts

  • Tensors
  • Autograd Engine

Building Models

  • Neural Networks
  • Optimizers

Hardware

  • GPU Acceleration

Documentation

Everything you need to build, train, and deploy deep learning models with RUMUS. From first principles to production-ready GPU-accelerated pipelines — all with the safety and performance guarantees of Rust.

Getting Started

Install RUMUS, build your first neural network, and train it in minutes. A hands-on introduction to the framework.

Read more

Tensors

Learn how RUMUS tensors work — creation, indexing, broadcasting, and efficient memory layout backed by Rust's ownership model.

Read more

Autograd

Understand automatic differentiation in RUMUS. Build computation graphs, call backward(), and inspect gradients with zero-cost abstractions.

Read more

Neural Networks

Compose models with Linear, Conv2d, MaxPool2d, Flatten, and Dropout layers. Use the #[derive(Module)] proc macro for ergonomic model definitions.

Read more

Optimizers

Train your models with SGD, Adam, and AdamW optimizers. Configure learning rates, momentum, weight decay, and scheduler strategies.

Read more

GPU Acceleration

Accelerate training and inference with WGPU. Move tensors to the GPU, run cross-platform compute shaders, and benchmark performance.

Read more
R
RUMUS

A native-Rust deep learning framework with memory safety, zero-cost abstractions, and GPU acceleration.

Framework

  • Get Started
  • Documentation
  • Community

Documentation

  • Tensors
  • Autograd
  • Neural Networks
  • Optimizers
  • GPU Acceleration

Resources

  • GitHub
  • Issues
  • Releases

RUMUS is developed by SUM INNOVATION INC Released under the MIT License