This repository contains the official project page for the paper:

POET-X: Memory-efficient LLM Training by Scaling Orthogonal Transformation (paper).

The implementation of POET-X is host at https://github.com/Sphere-AI-Lab/poet