CASP: Compression of Large Multimodal Models Based on Attention Sparsity

• URL: http://arxiv.org/abs/2503.05936v1
• Date: Fri, 07 Mar 2025 21:11:44 GMT
• Title: CASP: Compression of Large Multimodal Models Based on Attention Sparsity
• Authors: Mohsen Gholami, Mohammad Akbari, Kevin Cannons, Yong Zhang,
• Abstract summary: We introduce CASP, a model compression technique for Large Multimodal Models (LMMs)<n>Our approach performs a data-aware low-rank decomposition on the Query and Key weight matrix, followed by quantization across all layers based on an optimal bit allocation process.<n>It is compatible with any quantization technique and enhances state-of-the-art 2-bit quantization methods (AQLM and QuIP#) by an average of 21% on image- and video-language benchmarks.
• Score: 8.385413230868805
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: In this work, we propose an extreme compression technique for Large Multimodal Models (LMMs). While previous studies have explored quantization as an efficient post-training compression method for Large Language Models (LLMs), low-bit compression for multimodal models remains under-explored. The redundant nature of inputs in multimodal models results in a highly sparse attention matrix. We theoretically and experimentally demonstrate that the attention matrix's sparsity bounds the compression error of the Query and Key weight matrices. Based on this, we introduce CASP, a model compression technique for LMMs. Our approach performs a data-aware low-rank decomposition on the Query and Key weight matrix, followed by quantization across all layers based on an optimal bit allocation process. CASP is compatible with any quantization technique and enhances state-of-the-art 2-bit quantization methods (AQLM and QuIP#) by an average of 21% on image- and video-language benchmarks.

Related papers

• MGAA: Multi-Granular Adaptive Allocation fof Low-Rank Compression of LLMs [9.244526043014098]
  Multi-Granular Adaptive Allocation (MGAA) method can adaptively allocate parameters between and within sublayers without task-specific evaluations in the compression process.<n> Comprehensive evaluations of MGAA across multiple LLMs backbone models and benchmark datasets demonstrate its superior performance.
  arXiv  Detail & Related papers  (2025-07-04T04:54:01Z)
• Benchmarking Post-Training Quantization in LLMs: Comprehensive Taxonomy, Unified Evaluation, and Comparative Analysis [89.60263788590893]
  Post-training Quantization (PTQ) technique has been extensively adopted for large language models (LLMs) compression.<n>Existing algorithms focus primarily on performance, overlooking the trade-off among model size, performance, and quantization bitwidth.
  arXiv  Detail & Related papers  (2025-02-18T07:35:35Z)
• Pushing the Limits of Large Language Model Quantization via the Linearity Theorem [71.3332971315821]
  We present a "line theoremarity" establishing a direct relationship between the layer-wise $ell$ reconstruction error and the model perplexity increase due to quantization. This insight enables two novel applications: (1) a simple data-free LLM quantization method using Hadamard rotations and MSE-optimal grids, dubbed HIGGS, and (2) an optimal solution to the problem of finding non-uniform per-layer quantization levels.
  arXiv  Detail & Related papers  (2024-11-26T15:35:44Z)
• AsymKV: Enabling 1-Bit Quantization of KV Cache with Layer-Wise Asymmetric Quantization Configurations [36.63586957377984]
  Large language models often require substantial storage space. Due to their massive parameter count, these models often require substantial storage space. One research direction proposes to compress the models using integer replacements for floating-point numbers.
  arXiv  Detail & Related papers  (2024-10-17T04:35:57Z)
• Pruning via Merging: Compressing LLMs via Manifold Alignment Based Layer Merging [14.123313596780726]
  We propose Manifold-Based Knowledge Alignment and Layer Merging Compression (MKA) MKA uses manifold learning and the Normalized Pairwise Information Bottleneck measure to merge similar layers, reducing model size while preserving essential performance. Our findings show that MKA not only preserves model performance but also achieves substantial compression ratios, outperforming traditional pruning methods.
  arXiv  Detail & Related papers  (2024-06-24T05:57:55Z)
• 2DQuant: Low-bit Post-Training Quantization for Image Super-Resolution [83.09117439860607]
  Low-bit quantization has become widespread for compressing image super-resolution (SR) models for edge deployment. It is notorious that low-bit quantization degrades the accuracy of SR models compared to their full-precision (FP) counterparts. We present a dual-stage low-bit post-training quantization (PTQ) method for image super-resolution, namely 2DQuant, which achieves efficient and accurate SR under low-bit quantization.
  arXiv  Detail & Related papers  (2024-06-10T06:06:11Z)
• SliM-LLM: Salience-Driven Mixed-Precision Quantization for Large Language Models [63.118592279833656]
  Post-training quantization (PTQ) is an effective technique for compressing large language models (LLMs)<n>We propose SliM-LLM, a salience-driven mixed-precision quantization framework that allocates bit-widths at the group-wise.<n> Experiments show that SliM-LLM achieves superior performance across various LLMs at low bit-widths.
  arXiv  Detail & Related papers  (2024-05-23T16:21:48Z)
• Adaptive Feature-based Low-Rank Compression of Large Language Models via Bayesian Optimization [42.53133823994923]
  Low-rank compression is a promising technique to reduce non-essential parameters in large language models.<n>We conduct empirical research on the low-rank characteristics of large models.<n>We propose a low-rank compression method suitable for large language models.
  arXiv  Detail & Related papers  (2024-05-17T08:27:12Z)
• Quantization of Large Language Models with an Overdetermined Basis [73.79368761182998]
  We introduce an algorithm for data quantization based on the principles of Kashin representation. Our findings demonstrate that Kashin Quantization achieves competitive or superior quality in model performance.
  arXiv  Detail & Related papers  (2024-04-15T12:38:46Z)
• Data-free Weight Compress and Denoise for Large Language Models [96.68582094536032]
  We propose a novel approach termed Data-free Joint Rank-k Approximation for compressing the parameter matrices.<n>We achieve a model pruning of 80% parameters while retaining 93.43% of the original performance without any calibration data.
  arXiv  Detail & Related papers  (2024-02-26T05:51:47Z)
• Extreme Compression of Large Language Models via Additive Quantization [59.3122859349777]
  Our algorithm, called AQLM, generalizes the classic Additive Quantization (AQ) approach for information retrieval. We provide fast GPU and CPU implementations of AQLM for token generation, which enable us to match or outperform optimized FP16 implementations for speed.
  arXiv  Detail & Related papers  (2024-01-11T18:54:44Z)
• Low-Rank Prune-And-Factorize for Language Model Compression [18.088550230146247]
  Matrix factorization fails to retain satisfactory performance under moderate to high compression rate. We propose two techniques: sparsity-aware SVD and mixed-rank fine-tuning.
  arXiv  Detail & Related papers  (2023-06-25T07:38:43Z)
• OPQ: Compressing Deep Neural Networks with One-shot Pruning-Quantization [32.60139548889592]
  We propose a novel One-shot Pruning-Quantization (OPQ) in this paper. OPQ analytically solves the compression allocation with pre-trained weight parameters only. We propose a unified channel-wise quantization method that enforces all channels of each layer to share a common codebook.
  arXiv  Detail & Related papers  (2022-05-23T09:05:25Z)
• Compression of Generative Pre-trained Language Models via Quantization [62.80110048377957]
  We find that previous quantization methods fail on generative tasks due to the textithomogeneous word embeddings We propose a token-level contrastive distillation to learn distinguishable word embeddings, and a module-wise dynamic scaling to make quantizers adaptive to different modules.
  arXiv  Detail & Related papers  (2022-03-21T02:11:35Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.