Salvato in:
| Autori principali: | , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | https://arxiv.org/abs/2509.11177 |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866914039002562560 |
|---|---|
| author | Guo, Hang Li, Yawei Benini, Luca |
| author_facet | Guo, Hang Li, Yawei Benini, Luca |
| contents | Recent advances in Large Language Model (LLM) compression, such as quantization and pruning, have achieved notable success. However, as these techniques gradually approach their respective limits, relying on a single method for further compression has become increasingly challenging. In this work, we explore an alternative solution by combining quantization and sparsity. This joint approach, though promising, introduces new difficulties due to the inherently conflicting requirements on weight distributions: quantization favors compact ranges, while pruning benefits from high variance. To attack this problem, we propose Optimal Brain Restoration (OBR), a general and training-free framework that aligns pruning and quantization by error compensation between both. OBR minimizes performance degradation on downstream tasks by building on a second-order Hessian objective, which is then reformulated into a tractable problem through surrogate approximation and ultimately reaches a closed-form solution via group error compensation. Experiments show that OBR enables aggressive W4A4KV4 quantization with 50% sparsity on existing LLMs, and delivers up to 4.72x speedup and 6.4x memory reduction compared to the FP16-dense baseline. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_11177 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Optimal Brain Restoration for Joint Quantization and Sparsification of LLMs Guo, Hang Li, Yawei Benini, Luca Computation and Language Recent advances in Large Language Model (LLM) compression, such as quantization and pruning, have achieved notable success. However, as these techniques gradually approach their respective limits, relying on a single method for further compression has become increasingly challenging. In this work, we explore an alternative solution by combining quantization and sparsity. This joint approach, though promising, introduces new difficulties due to the inherently conflicting requirements on weight distributions: quantization favors compact ranges, while pruning benefits from high variance. To attack this problem, we propose Optimal Brain Restoration (OBR), a general and training-free framework that aligns pruning and quantization by error compensation between both. OBR minimizes performance degradation on downstream tasks by building on a second-order Hessian objective, which is then reformulated into a tractable problem through surrogate approximation and ultimately reaches a closed-form solution via group error compensation. Experiments show that OBR enables aggressive W4A4KV4 quantization with 50% sparsity on existing LLMs, and delivers up to 4.72x speedup and 6.4x memory reduction compared to the FP16-dense baseline. |
| title | Optimal Brain Restoration for Joint Quantization and Sparsification of LLMs |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2509.11177 |