mni-ml/framework

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

1010 stars · 130 forks observed · 2026-09-22

What it is AI-extracted, prompt v1, taxonomy v1, 2026-09-07, confidence 0.9

A TypeScript machine learning framework with Rust native backends for CPU, CUDA, and WebGPU. It provides autograd, tensor operations, built-in neural network modules, and optimizers with a PyTorch-like API running at GPU speed.

Use cases

• train neural networks from TypeScript
• run GPU-accelerated tensor math with CUDA or WebGPU
• learn how ML frameworks and autograd work internally
• build models with a PyTorch-like API in JavaScript/TypeScript
• train convnets and MLPs in the browser via WebGPU
• experiment with deep learning without leaving the Node ecosystem

When to choose

• you want to write ML training code in TypeScript with native GPU acceleration
• you need WebGPU support for browser or cross-vendor GPU compute
• you want a readable, hackable PyTorch-like framework to study internals

When to avoid

• you need production-scale training with mature ecosystem support like PyTorch or JAX
• you require extensive pretrained model hubs or a large op coverage
• your project depends on a battle-tested framework with long-term stability guarantees

Facets

library · maturity active

machine-learning deep-learning llm-training machine-learning deep-learning developer-tools rust cross-platform autograd tensor-operations cuda-kernels pytorch-like-api neural-networks webgpu typescript gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="mni-ml/framework")