float8

PyTorch native quantization and sparsity for training and inference

A modular, accelerator-ready machine learning framework built in Go that speaks float8/16/32/64. Designed with clean architecture, strong typing, and native concurrency for scalable, production-ready AI systems. Ideal for engineers who value simplicity, speed, and maintainability.

A library written in C for converting between float8 (8-bit minifloat numbers) and float32 (single-precision floating-point numbers) formats.

minifloat (8-bit float) in Golang

Official Code for the paper ELMO : Efficiency via Low-precision and Peak Memory Optimization in Large Output Spaces (in ICML 2025)