Results for supervised_learning

A Machine Learning framework for Rust

Safe Rust. No Deps. RL for the CPU.

Semi-supervised learning algorithms

Zero-config nonparametric supervised learning — KNN with auto-tuned K, metric learning, and VP-tree indexing

Checkpoint and ONNX compatibility utilities for jepa-rs models

Core traits and abstractions for JEPA (Joint Embedding Predictive Architecture)

Training loop utilities for JEPA models

Vision-specific JEPA implementations (ViT encoder, patch embedding, I-JEPA)

Action-conditioned world model and hierarchical JEPA (H-JEPA)

A collection of tree-based algorithms

Self-supervised learning primitives for OxiCUDA: SimCLR, MoCo, BYOL, Barlow Twins, VICReg, MAE, SwAV, DINO — pure Rust, zero CUDA SDK dependency.

CLI and TUI for the jepa-rs workspace