Quartet II: Accurate LLM Pre-Training in NVFP4 by Improved Unbiased Gradient Estimation

Andrei Panferov, Erik Schultheis, Soroush Tabesh, Dan Alistarh

TL;DR: We provide a new method for training in the NVFP4 format by leveraging a new unbiased gradient estimator.

Abstract:

The NVFP4 lower-precision format, supported in hardware by NVIDIA Blackwell GPUs, promises to allow, for the first time, end-to-end fully-quantized pre-training of massive models such as LLMs.
Yet, existing quantized training methods still sacrifice some of the representation capacity of this format in favor of more accurate unbiased quantized gradient estimation by stochastic rounding (SR), losing noticeable accuracy relative to standard FP16 and FP8 training.
In this paper, improve the state of the art for quantized training in NVFP4 via a novel unbiased quantization routine for micro-scaled formats, called MS-EDEN, that has more than 2x lower quantization error than SR. We integrate it into a novel fully-NVFP4 quantization scheme for linear layers, called Quartet II. We show analytically that Quartet II achieves consistently better gradient estimation across all major matrix multiplications, both on the forward and on the backward passes.
In addition, our proposal synergizes well with recent training improvements aimed specifically at NVFP4. We further validate Quartet II on end-to-end LLM training with up to 1.9B parameters on 38B tokens. We provide kernels for execution on NVIDIA Blackwell GPUs with up to 4.2x speedup over BF16.

Lay Summary: NVFP4 is the latest computation format introduced by NVIDIA that greatly speeds up matrix operations computations. We design a novel method for pre-training LLMs with majority of computations in NVFP4 that improves model quality via more accurate compression schemes.

Link To Code: https://github.com/IST-DASLab/Quartet-II/

Primary Area: Deep Learning->Algorithms

Keywords: Quantization, Training, Gradient Estimators, GPU Kernels

Submission Number: 32248