A Training-free Sub-quadratic Cost Transformer Model Serving Framework With Hierarchically Pruned Attention
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866929684300693504 |
|---|---|
| author | Lee, Heejun Park, Geon Lee, Youngwan Suh, Jaduk Kim, Jina Jeong, Wonyoung Kim, Bumsik Lee, Hyemin Jeon, Myeongjae Hwang, Sung Ju |
| author_facet | Lee, Heejun Park, Geon Lee, Youngwan Suh, Jaduk Kim, Jina Jeong, Wonyoung Kim, Bumsik Lee, Hyemin Jeon, Myeongjae Hwang, Sung Ju |
| contents | In modern large language models (LLMs), increasing the context length is crucial for improving comprehension and coherence in long-context, multi-modal, and retrieval-augmented language generation. While many recent transformer models attempt to extend their context length over a million tokens, they remain impractical due to the quadratic time and space complexities. Although recent works on linear and sparse attention mechanisms can achieve this goal, their real-world applicability is often limited by the need to re-train from scratch and significantly worse performance. In response, we propose a novel approach, Hierarchically Pruned Attention (HiP), which reduces the time complexity of the attention mechanism to $O(T \log T)$ and the space complexity to $O(T)$, where $T$ is the sequence length. We notice a pattern in the attention scores of pretrained LLMs where tokens close together tend to have similar scores, which we call ``attention locality''. Based on this observation, we utilize a novel tree-search-like algorithm that estimates the top-$k$ key tokens for a given query on the fly, which is mathematically guaranteed to have better performance than random attention pruning. In addition to improving the time complexity of the attention mechanism, we further optimize GPU memory usage by implementing KV cache offloading, which stores only $O(\log T)$ tokens on the GPU while maintaining similar decoding throughput. Experiments on benchmarks show that HiP, with its training-free nature, significantly reduces both prefill and decoding latencies, as well as memory usage, while maintaining high-quality generation with minimal degradation. HiP enables pretrained LLMs to scale up to millions of tokens on commodity GPUs, potentially unlocking long-context LLM applications previously deemed infeasible. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2406_09827 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | A Training-free Sub-quadratic Cost Transformer Model Serving Framework With Hierarchically Pruned Attention Lee, Heejun Park, Geon Lee, Youngwan Suh, Jaduk Kim, Jina Jeong, Wonyoung Kim, Bumsik Lee, Hyemin Jeon, Myeongjae Hwang, Sung Ju Computation and Language Computer Vision and Pattern Recognition Distributed, Parallel, and Cluster Computing Machine Learning In modern large language models (LLMs), increasing the context length is crucial for improving comprehension and coherence in long-context, multi-modal, and retrieval-augmented language generation. While many recent transformer models attempt to extend their context length over a million tokens, they remain impractical due to the quadratic time and space complexities. Although recent works on linear and sparse attention mechanisms can achieve this goal, their real-world applicability is often limited by the need to re-train from scratch and significantly worse performance. In response, we propose a novel approach, Hierarchically Pruned Attention (HiP), which reduces the time complexity of the attention mechanism to $O(T \log T)$ and the space complexity to $O(T)$, where $T$ is the sequence length. We notice a pattern in the attention scores of pretrained LLMs where tokens close together tend to have similar scores, which we call ``attention locality''. Based on this observation, we utilize a novel tree-search-like algorithm that estimates the top-$k$ key tokens for a given query on the fly, which is mathematically guaranteed to have better performance than random attention pruning. In addition to improving the time complexity of the attention mechanism, we further optimize GPU memory usage by implementing KV cache offloading, which stores only $O(\log T)$ tokens on the GPU while maintaining similar decoding throughput. Experiments on benchmarks show that HiP, with its training-free nature, significantly reduces both prefill and decoding latencies, as well as memory usage, while maintaining high-quality generation with minimal degradation. HiP enables pretrained LLMs to scale up to millions of tokens on commodity GPUs, potentially unlocking long-context LLM applications previously deemed infeasible. |
| title | A Training-free Sub-quadratic Cost Transformer Model Serving Framework With Hierarchically Pruned Attention |
| topic | Computation and Language Computer Vision and Pattern Recognition Distributed, Parallel, and Cluster Computing Machine Learning |
| url | https://arxiv.org/abs/2406.09827 |