paper-with-me

홈 › Papers

Max-Window Scale Estimation for Near-Lossless HiF8 W8A8 Quantization-Aware Training

2026-05-25 · Yingying Cheng, Jinquan Shi, Li Zhou, Zhiyang He, Zhaoyi Sun, Fan Zhang, Jie Sun arxiv

Quantization-aware training (QAT) with low-bit floating-point formats enables efficient LLM deployment, yet introduces subtle failure modes invisible to standard training metrics. We present a systematic study of HiF8 W8A8 QAT for OpenPangu-Embedded-1B through the lens of Delayed Tensor Scaling (DTS). Across eight controlled experiments, we identify and disentangle two orthogonal failure modes: (i)amax saturation, where delayed scale estimates silently corrupt knowledge-sensitive representations via forward-pass clipping, and (ii)catastrophic forgetting, where an aggressive learning rate overwrites pretrained commonsense knowledge independently of quantization. Neither is detectable from training loss alone. We address amax saturation with a conservative max-algorithm DTS strategy over a 64-step history window, and mitigate forgetting via a 500-step BF16 warmup followed by QAT at lr=10^{-5}. Both fixes are necessary and sufficient: our final configuration achieves 0.43% MMLU drop, 0.58% HellaSwag drop, and 0.22% ARC-Challenge drop versus a matched BF16 baseline, with a training loss APE of only 0.11% over 10,000 steps.

📄 PDF Abstract BibTeX arXiv:2605.26189

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

LQER: Low-Rank Quantization Error Reconstruction for LLMs

2024-02-04 · Cheng Zhang, Jianyi Cheng, George A. Constantinides, Yiren Zhao

Post-training quantization of Large Language Models (LLMs) is challenging. In this work, we introduce Low-rank Quantization Error Reduction (LQER), which combines quantization and low-rank approximation to recover the mo…

Knowledge DistillationQuantization

HyQuant: Hybrid-Precision Quantization for LLM Attention

2026-08-28 · Jiatong Ding, Bingxin Xing, Yu Zhang, Dian Ding 외 arxiv

Quantization has been widely adopted in LLM training and inference to reduce cost and improve efficiency. However, low-bit quantization of the attention module often introduces large errors at very low bit-widths, causin…

MixA-Q: Revisiting Activation Sparsity for Vision Transformers from a Mixed-Precision Quantization Perspective

2025-07-25 · Weitian Wang, Rai Shubham, Cecilia De La Parra, Akash Kumar arxiv

In this paper, we propose MixA-Q, a mixed-precision activation quantization framework that leverages intra-layer activation sparsity (a concept widely explored in activation pruning methods) for efficient inference of qu…

Lossless Compression of Vector IDs for Approximate Nearest Neighbor Search

2025-01-16 · Daniel Severo, Giuseppe Ottaviano, Matthew Muckley, Karen Ullrich 외

Approximate nearest neighbor search for vectors relies on indexes that are most often accessed from RAM. Therefore, storage is the factor limiting the size of the database that can be served from a machine. Lossy vector …

Quantization

Kirin: Improving ANN efficiency with SNN Hybridization

2026-02-09 · Chenyu Wang, Zhanglu Yan, Zhi Zhou, Xu Chen 외 arxiv

Artificial neural networks (ANNs), particularly large language models (LLMs), demonstrate powerful inference capabilities but consume substantial energy. Conversely, spiking neural networks (SNNs) exhibit exceptional ene…