LiveMind: Low-latency Large Language Models with Simultaneous Inference

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Chen, Chuangtao, Zhang, Grace Li, Yin, Xunzhao, Zhuo, Cheng, Schlichtmann, Ulf, Li, Bing
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915006754324480
author Chen, Chuangtao
Zhang, Grace Li
Yin, Xunzhao
Zhuo, Cheng
Schlichtmann, Ulf
Li, Bing
author_facet Chen, Chuangtao
Zhang, Grace Li
Yin, Xunzhao
Zhuo, Cheng
Schlichtmann, Ulf
Li, Bing
contents In this paper, we introduce LiveMind, a novel low-latency inference framework for large language model (LLM) inference which enables LLMs to perform inferences with incomplete user input. By reallocating computational processes to the input phase, a substantial reduction in latency is achieved, thereby significantly enhancing the interactive experience for users of LLMs. The framework adeptly manages the visibility of the streaming input to the model, allowing it to infer from incomplete user input or await additional content. Compared with traditional inference methods on complete user input, our approach demonstrates an average reduction in response latency of 84.0% on the MMLU dataset and 71.6% on the MMLU-Pro dataset, while maintaining comparable accuracy. Additionally, our framework facilitates collaborative inference and output across different models. By employing an large LLM for inference and a small LLM for output, we achieve an average 37% reduction in response latency, alongside a 4.30% improvement in accuracy on the MMLU-Pro dataset compared with the baseline. The proposed LiveMind framework advances the field of human-AI interaction by enabling more responsive and efficient communication between users and AI systems.
format Preprint
id arxiv_https___arxiv_org_abs_2406_14319
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle LiveMind: Low-latency Large Language Models with Simultaneous Inference
Chen, Chuangtao
Zhang, Grace Li
Yin, Xunzhao
Zhuo, Cheng
Schlichtmann, Ulf
Li, Bing
Artificial Intelligence
Computation and Language
In this paper, we introduce LiveMind, a novel low-latency inference framework for large language model (LLM) inference which enables LLMs to perform inferences with incomplete user input. By reallocating computational processes to the input phase, a substantial reduction in latency is achieved, thereby significantly enhancing the interactive experience for users of LLMs. The framework adeptly manages the visibility of the streaming input to the model, allowing it to infer from incomplete user input or await additional content. Compared with traditional inference methods on complete user input, our approach demonstrates an average reduction in response latency of 84.0% on the MMLU dataset and 71.6% on the MMLU-Pro dataset, while maintaining comparable accuracy. Additionally, our framework facilitates collaborative inference and output across different models. By employing an large LLM for inference and a small LLM for output, we achieve an average 37% reduction in response latency, alongside a 4.30% improvement in accuracy on the MMLU-Pro dataset compared with the baseline. The proposed LiveMind framework advances the field of human-AI interaction by enabling more responsive and efficient communication between users and AI systems.
title LiveMind: Low-latency Large Language Models with Simultaneous Inference
topic Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2406.14319