Metraipolis › archive › dc_0feb4e1b9219
AutoCompress: Automated Hardware-Guided Model Compression — arXiv:2611.08290
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.08290
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-06T14:42:53.768Z
- SHA-256 of the archived text
89748f823d3bd9f17f06acdd0972dc4a8287d34009c46b452c908cff4e2d7068- Stable URL of this page
https://metraipolis.com/documento/dc_0feb4e1b9219
How to cite
AutoCompress: Automated Hardware-Guided Model Compression — arXiv:2611.08290 — arXiv (en), https://arxiv.org/abs/2611.08290 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_0feb4e1b9219 (Metraipolis archive dc_0feb4e1b9219, sha256 89748f823d3bd9f17f06acdd0972dc4a8287d34009c46b452c908cff4e2d7068)
BibTeX
@misc{dc0feb4e1b9219,
title = {AutoCompress: Automated Hardware-Guided Model Compression — arXiv:2611.08290},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.08290},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_0feb4e1b9219; sha256 89748f823d3bd9f17f06acdd0972dc4a8287d34009c46b452c908cff4e2d7068},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.08290. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
LLM pruning must preserve quality while meeting hardware- and workload-specific constraints on latency, throughput, and memory. Existing pruning workflows often target parameter counts, requiring practitioners to manually select model sizes that meet latency, throughput, and GPU memory limits. However, parameter count alone does not predict these metrics, which depend on the model architecture, workload, hardware, and serving engine. We introduce AutoCompress, a hardware-guided compression framework that directly incorporates such constraints into differentiable architecture search. AutoCompress jointly optimizes global and per-layer structure to produce heterogeneous architectures, guided by serving-engine cost models validated against end-to-end latency, throughput, and memory measurements. We evaluate AutoCompress on long-context local coding, batched datacenter serving, and latency-sensitive in-game interaction. Compared to the state-of-the-art Minitron pruning framework, AutoCompress improves average accuracy across eight benchmarks by 6.57, 2.35, and 0.57 percentage points on RTX 5070, RTX 5080, and H100, respectively. AutoCompress also compresses Nemotron 3.5 Super from 120B to 74B parameters for single-B200 BF16 serving at 128K context, retaining 95% of its mean benchmark score. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2611.08290 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-11-06T14:18:42.700Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).