In-context KV-Cache Eviction for LLMs via Attention-Gate

• URL: http://arxiv.org/abs/2410.12876v2
• Date: Sat, 19 Oct 2024 08:45:11 GMT
• Title: In-context KV-Cache Eviction for LLMs via Attention-Gate
• Authors: Zihao Zeng, Bokai Lin, Tianqi Hou, Hao Zhang, Zhijie Deng,
• Abstract summary: The KV-Cache technique has become the standard for the inference of large language models (LLMs) This paper devises a parameterized KV-Cache eviction mechanism, dubbed as Attention-Gate. Attention-Gate accepts the whole context as input and yields eviction flags for each token to realize in-context eviction.
• Score: 12.732519329131392
• License:
• Abstract: The KV-Cache technique has become the standard for the inference of large language models (LLMs). It caches states of self-attention to avoid recomputation. Yet, it is widely criticized that KV-Cache can become a bottleneck of the LLM inference system, especially when confronted with ultra-large models and long-context queries. A natural remedy is to discard the KV-Cache for less important tokens, with StreamingLLM as an example, but the used static eviction strategies cannot flexibly adapt to varying contexts. Remedies like H2O leverage accumulative attention scores to perform dynamic eviction but suffer from the attention bias issue in capturing contextual information. This paper bridges this gap by devising a parameterized KV-Cache eviction mechanism, dubbed as Attention-Gate, which accepts the whole context as input and yields eviction flags for each token to realize in-context eviction. The subsequent self-attention module proceeds according to the flags and only the KV states for the remaining tokens need to be cached. The Attention-Gates can vary among different heads and layers and be trivially plugged into pre-trained LLMs, tuned by cost-effective continual pre-training or supervised fine-tuning objectives to acquire what to discard. The computational and memory overhead introduced by Attention-Gates is minimal. Our method is validated across multiple tasks, demonstrating both efficiency and adaptability. After a highly efficient continual pre-training, it achieves higher average accuracy and evicts more tokens compared to traditional training-free methods. In supervised fine-tuning, it not only evicts many tokens but also outperforms LoRA-finetuned LLMs on some datasets, such as RTE, where it improves accuracy by 13.9% while evicting 62.8% of tokens, showing that effective eviction of redundant tokens can even enhance performance.

Related papers

• NACL: A General and Effective KV Cache Eviction Framework for LLMs at Inference Time [44.89402186438295]
  Large Language Models (LLMs) have ignited an innovative surge of AI applications, marking a new era of exciting possibilities equipped with extended context windows. However, hosting these models is cost-prohibitive mainly due to the extensive memory consumption of KV Cache involving long-context modeling. We propose NACL, a general framework for long-context KV cache eviction that achieves more optimal and efficient eviction in a single operation during the encoding phase.
  arXiv  Detail & Related papers  (2024-08-07T10:31:07Z)
• ThinK: Thinner Key Cache by Query-Driven Pruning [63.13363917871414]
  Large Language Models (LLMs) have revolutionized the field of natural language processing, achieving unprecedented performance across a variety of applications. This paper focuses on the long-context scenario, addressing the inefficiencies in KV cache memory consumption during inference. We propose ThinK, a novel query-dependent KV cache pruning method designed to minimize attention weight loss while selectively pruning the least significant channels.
  arXiv  Detail & Related papers  (2024-07-30T17:59:08Z)
• Efficient Inference of Vision Instruction-Following Models with Elastic Cache [76.44955111634545]
  We introduce Elastic Cache, a novel strategy for efficient deployment of instruction-following large vision-language models. We propose an importance-driven cache merging strategy to prune redundancy caches. For instruction encoding, we utilize the frequency to evaluate the importance of caches. Results on a range of LVLMs demonstrate that Elastic Cache not only boosts efficiency but also notably outperforms existing pruning methods in language generation.
  arXiv  Detail & Related papers  (2024-07-25T15:29:05Z)
• Ada-KV: Optimizing KV Cache Eviction by Adaptive Budget Allocation for Efficient LLM Inference [19.447729423696096]
  Large Language Models have excelled in various fields but encounter challenges in memory and time efficiency. Recent efforts try to reduce KV cache size to a given memory budget by evicting vast non-critical cache elements during runtime.
  arXiv  Detail & Related papers  (2024-07-16T09:53:32Z)
• D2O: Dynamic Discriminative Operations for Efficient Generative Inference of Large Language Models [14.665924387149014]
  Efficient inference in Large Language Models (LLMs) is impeded by the growing memory demands of key-value (KV) caching. Traditional KV cache eviction strategies prioritize less critical KV-pairs based on attention scores, leading to issues such as context loss or hallucinations. We introduce Dynamic Discriminative Operations (D2O), a novel method that utilizes two-level discriminative strategies to optimize KV cache size without fine-tuning.
  arXiv  Detail & Related papers  (2024-06-18T20:01:51Z)
• FFN-SkipLLM: A Hidden Gem for Autoregressive Decoding with Adaptive Feed Forward Skipping [49.66872823080736]
  Autoregressive Large Language Models (e.g., LLaMa, GPTs) are omnipresent achieving remarkable success in language understanding and generation. To mitigate overload incurred during generation, several early-exit and layer-dropping strategies have been proposed. We propose FFN-SkipLLM, which is an input-adaptive feed-forward skipping strategy.
  arXiv  Detail & Related papers  (2024-04-05T02:35:43Z)
• QAQ: Quality Adaptive Quantization for LLM KV Cache [3.163526369095745]
  A bottleneck in model deployment emerges due to the linear expansion of the Key-Value cache with the context length. We propose QAQ, a Quality Adaptive Quantization scheme for the KV cache.
  arXiv  Detail & Related papers  (2024-03-07T16:42:37Z)
• Get More with LESS: Synthesizing Recurrence with KV Cache Compression for Efficient LLM Inference [78.65321721142624]
  We focus on a memory bottleneck imposed by the key-value ( KV) cache. Existing KV cache methods approach this problem by pruning or evicting large swaths of relatively less important KV pairs. We propose LESS, a simple integration of a constant sized cache with eviction-based cache methods.
  arXiv  Detail & Related papers  (2024-02-14T18:54:56Z)
• Model Tells You What to Discard: Adaptive KV Cache Compression for LLMs [82.08922896531618]
  We introduce adaptive KV cache compression, a plug-and-play method that reduces the memory footprint of generative inference for Large Language Models (LLMs) We conduct targeted profiling to discern the intrinsic structure of attention modules. Based on the recognized structure, we then construct the KV cache in an adaptive manner: evicting long-range contexts on attention heads emphasizing local contexts, discarding non-special tokens on attention heads centered on special tokens, and only employing the standard KV cache for attention heads that broadly attend to all tokens.
  arXiv  Detail & Related papers  (2023-10-03T05:17:08Z)

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.