ZACK: Zero-Overhead LLM Inference Acceleration via Dimensionality Compression of the Key-Value Cache

• URL: http://arxiv.org/abs/2408.04107v2
• Date: Wed, 05 Feb 2025 21:44:56 GMT
• Title: ZACK: Zero-Overhead LLM Inference Acceleration via Dimensionality Compression of the Key-Value Cache
• Authors: Zeyu Zhang, Haiying Shen,
• Abstract summary: We propose ZACK, the first KV dimensionality compression system that achieves zero-overhead compression and decompression and also reduces attention time. ZACK employs adaptive compression, tailoring KV compression rates across heads and layers based on their contributions to inference. Comprehensive experiments demonstrate that when combined with ZACK, state-of-the-art eviction-based and quantization-based methods for KV compression further reduce KV size by up to 68%, Time-To-First-Token (TTFT) by up to 44%, and Time-Between-Tokens (TBT) by up to 55%.
• Score: 11.194752361478567
• License:
• Abstract: In large-language models, memory constraints in the Key-Value Cache (KVC) pose a challenge during inference. In this work, we propose ZACK, the first KV dimensionality compression system that achieves zero-overhead compression and decompression and also reduces attention computation time. It complements and can be combined with eviction-based and quantization-based methods to further enhance KV compression. Moreover, ZACK employs adaptive compression, tailoring KV compression rates across heads and layers based on their contributions to inference to maximize overall compression while maintaining an accuracy loss constraint. Additionally, ZACK enhances the self-attention kernel to balance the uneven workloads caused by the adaptive compression approach to further reduce attention computation latency. Comprehensive experiments demonstrate that when combined with ZACK, state-of-the-art eviction-based and quantization-based methods for KV compression further reduce KV size by up to 68%, Time-To-First-Token (TTFT) by up to 44%, and Time-Between-Tokens (TBT) by up to 55% and achieve up to 1.72X throughput under the same latency, while maintaining 99% of the baseline accuracy. We open-sourced the code.

Related papers

• HACK: Homomorphic Acceleration via Compression of the Key-Value Cache for Disaggregated LLM Inference [24.068304021577358]
  Disaggregated Large Language Model (LLM) inference separates computation-intensive prefill stage from memory-intensive decode stage. transmitting Key-Value (KV) data between the two stages can be a bottleneck, especially for long prompts. We propose Homomorphic Acceleration via Compression of the KV cache (HACK) for disaggregated LLM inference.
  arXiv  Detail & Related papers  (2025-02-05T20:09:51Z)
• More Tokens, Lower Precision: Towards the Optimal Token-Precision Trade-off in KV Cache Compression [71.42818367729573]
  In large language models (LLMs), the memory usage of KV cache has become a critical bottleneck during inference. The mainstream KV compression methods, including KV pruning and KV quantization, primarily focus on either token or precision dimension separately. In this paper, we comprehensively investigate the token-precision trade-off in KV cache compression.
  arXiv  Detail & Related papers  (2024-12-17T09:20:31Z)
• EMS: Adaptive Evict-then-Merge Strategy for Head-wise KV Cache Compression Based on Global-Local Importance [44.14919492126948]
  As memory overhead becomes a significant concern, efficient compression of KV cache has gained increasing attention. We propose EMS to overcome these limitations, while achieving better KV cache compression under extreme compression ratios. EMS consistently achieves the lowest perplexity, improves scores by over 1.28 points across four LLMs on LongBench under a 256 cache budget, and preserves 95% retrieval accuracy with a cache budget less than 2% of the context length in the Needle-in-a-Haystack task.
  arXiv  Detail & Related papers  (2024-12-11T16:35:13Z)
• LoRC: Low-Rank Compression for LLMs KV Cache with a Progressive Compression Strategy [59.1298692559785]
  Key-Value ( KV) cache is crucial component in serving transformer-based autoregressive large language models (LLMs) Existing approaches to mitigate this issue include: (1) efficient attention variants integrated in upcycling stages; (2) KV cache compression at test time; and (3) KV cache compression at test time. We propose a low-rank approximation of KV weight matrices, allowing plug-in integration with existing transformer-based LLMs without model retraining. Our method is designed to function without model tuning in upcycling stages or task-specific profiling in test stages.
  arXiv  Detail & Related papers  (2024-10-04T03:10:53Z)
• UNComp: Uncertainty-Aware Long-Context Compressor for Efficient Large Language Model Inference [38.11539884622708]
  UNComp is an uncertainty-aware compression scheme that adaptively compresses both the hidden states and the KV cache. Our method achieves a 1.6x speedup in the prefilling stage and reduces the KV cache to 4.74% of its original size. Remarkably, in needle-in-a-haystack tasks, UNComp outperforms the full-size KV cache even when compressed to 9.38% of its original size.
  arXiv  Detail & Related papers  (2024-10-04T02:32:36Z)
• KV-Compress: Paged KV-Cache Compression with Variable Compression Rates per Attention Head [0.8158530638728501]
  We introduce KV-Compress, a novel compression method that evicts contiguous KV blocks within a PagedAttention framework. Our method achieves state-of-the-art performance on LongBench for both Mistral-7B-Instruct-v0.2 and Llama-3.1-8B-Instruct while lowering the total number of compressed KVs by 4x. Evaluations on Llama-3.1-8B-Instruct and Llama-3.1-70B-Instruct-FP8 achieve compression rates up to 8x with negligible impact on performance, and up to 64x while retaining over 90% of full-cache performance
  arXiv  Detail & Related papers  (2024-09-30T19:09:13Z)
• LoCoCo: Dropping In Convolutions for Long Context Compression [77.26610232994508]
  This paper presents a novel approach, Dropping In Convolutions for Long Context Compression (LoCoCo) LoCoCo employs only a fixed-size Key-Value ( KV) cache, and can enhance efficiency in both inference and fine-tuning stages.
  arXiv  Detail & Related papers  (2024-06-08T01:35:11Z)
• PyramidKV: Dynamic KV Cache Compression based on Pyramidal Information Funneling [53.08975547824068]
  We investigate whether attention-based information flow inside large language models (LLMs) is aggregated through noticeable patterns for long context processing. Our observations reveal that LLMs aggregate information through Pyramidal Information Funneling where attention is scattering widely in lower layers. Motivated by these insights, we developed Pyramid KV, a novel and effective KV cache compression method.
  arXiv  Detail & Related papers  (2024-06-04T07:51:30Z)
• ZipCache: Accurate and Efficient KV Cache Quantization with Salient Token Identification [19.985314022860432]
  KV cache stores key and value states from previous tokens to avoid re-computation. KV cache compression seeks to discern the saliency of tokens, preserving vital information while aggressively compressing those of less importance. We present ZipCache, an accurate and efficient KV cache quantization method for LLMs.
  arXiv  Detail & Related papers  (2024-05-23T07:37:16Z)
• Unlocking Data-free Low-bit Quantization with Matrix Decomposition for KV Cache Compression [87.5604418100301]
  Key-value( KV) caching is an important technique to accelerate the inference of large language models. Existing methods often compromise precision or require extra data for calibration. We introduce textbfDecoQuant, a novel data-free low-bit quantization technique based on tensor decomposition methods.
  arXiv  Detail & Related papers  (2024-05-21T08:35:10Z)
• Towards Compact CNNs via Collaborative Compression [166.86915086497433]
  We propose a Collaborative Compression scheme, which joints channel pruning and tensor decomposition to compress CNN models. We achieve 52.9% FLOPs reduction by removing 48.4% parameters on ResNet-50 with only a Top-1 accuracy drop of 0.56% on ImageNet 2012.
  arXiv  Detail & Related papers  (2021-05-24T12:07:38Z)

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.