arXiv：cs.LG· Ananyo Bhattacharya, Swastik Bhattacharya, Christiane Jablonowski·· 3 小时前AI 评分30
自回归天气模型的训练后量化研究
Post-Training Quantization of Autoregressive Weather Models
AI 导读
研究将训练后量化（PTQ）应用于深度学习天气预测（DLWP）和 FourCastNet（FCN）两个预训练全球天气预测模型，系统考察 PTQ 对自回归推理在短期预报时效内的影响。模拟量化配置的评估显示，短期时效内仍可给出定性上有意义的预报。该工作给出了自回归天气模拟器 PTQ 的首个基准。
正文
Abstract:Advancements in high-resolution numerical weather prediction (NWP) and data assimilation (DA) have shaped the developments in deep learning (DL) architectures emulating atmospheric dynamics. Emulators for weather forecasting exhibit forecast quality comparable to physics based models at forecast horizon scaling from few days to subseasonal time scales. The emulators are driven by hardware-accelerated matrix multiplication in autoregressive inferences, significantly reducing the computation time and resources required for NWP. Optimization of the matrix multiplication processes in GPU architectures provides opportunities to scale towards high-resolution domain, and offers implementation of out of the box solutions. Post-training quantization (PTQ) has been demonstrated across multiple DL architectures to accelerate and increase the number of computations in unit time while consuming less power, enabling applications on edge hardware. In this study, we investigate the effect of PTQ on pre-trained AI emulators for global-scale weather forecasting. We implement PTQ algorithms in Deep Learning Weather Prediction (DLWP) and FourCastNet (FCN) models as a proof of concept for geophysical fluid dynamics applications. We systematically investigate the effect of PTQ on emulator inferences over short-range forecast horizons. Evaluation of PTQ configurations using simulated quantization hints at qualitatively meaningful forecasts over short-time horizons. These results provide a first benchmark of PTQ for autoregressive weather emulators and a basis for quantization-based optimization of DL models for dynamical systems.
| Subjects: | Machine Learning (cs.LG); Earth and Planetary Astrophysics (astro-ph.EP) |
| Cite as: | arXiv:2610.02511 [cs.LG] |
| (or arXiv:2610.02511v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.02511 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Ananyo Bhattacharya [view email]
[v1]
Thu, 1 Oct 2026 21:34:27 UTC (14,697 KB)
来源：arXiv：cs.LG · arxiv.org