2112.10769

arxiv.org · 7,531 words · saved by 1 readers

N/A

Published as a conference paper at ICLR 2023 ACCURATE N EURAL T RAINING WITH 4- BIT M ATRIX M ULTIPLICATIONS AT S TANDARD F ORMATS Brian Chmiel †◦ Ron Banner † Elad Hoffer † Hilla Ben Yaacov † Daniel Soudry ◦ † Habana Labs – An Intel company, Caesarea, Israel, ◦ Department of…

saved by

Sankeerth Durvasula

related reading

lec06.pdfdropbox.com
[1706.04599] On Calibration of Modern Neural Networksarxiv.org
The 4-bitter Lesson | humans&humansand.ai
[2106.08295] A White Paper on Neural Network Quantizationarxiv.org
2502.11089arxiv.org
[2510.11696] QeRL: Beyond Efficiency -- Quantization-enhanced Reinforcement Learning for LLMsarxiv.org
[2309.10668] Language Modeling Is Compressionarxiv.org
[2510.13786] The Art of Scaling Reinforcement Learning Compute for LLMsarxiv.org
[2605.01172] A Theory of Generalization in Deep Learningarxiv.org
Integer Quantization for Deep Learning Inference: Principles and Empirical Evaluationarxiv.org
Understanding deep learning requires rethinking generalizationarxiv.org
arxiv.org/pdf/1805.08522arxiv.org