Context lengths of Large Language Models (LLMs) have exploded in recent years, with 128k-token context becoming a standard and million-token context becoming a reality. Efficiently supporting long-context inference remains challenging as the memory that must be allocated in key-value (KV) cache for a generation scales with its context length, limiting the number of long-context requests that can be served concurrently under a given memory budget. KV cache compression can mitigate this issue by removing under-utilized KVs from each attention head's cache and reducing its memory footprint. Higher theoretical compression rates can be achieved when the number of removed KVs varies across attention heads, but application of such a strategy within existing inference frameworks adds fragmentation and cannot realize the theoretical compression rates in physical memory. We introduce KV-Compress, a novel compression method that evicts contiguous KV blocks within a PagedAttention framework, reducing the memory footprint of the KV cache proportionally to this theoretical compression rate. 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 compared with prior methods. 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 for all but three of the suite's subsets. We benchmark an integration of our method with vLLM that increases total throughput by up to 5.18x by enabling larger decoding batches.
相關內容
By augmenting Large Language Models (LLMs) with external tools, their capacity to solve complex problems has been significantly enhanced. However, despite ongoing advancements in the parsing capabilities of LLMs, incorporating all available tools simultaneously in the prompt remains impractical due to the vast number of external tools. Consequently, it is essential to provide LLMs with a precise set of tools tailored to the specific task, considering both quantity and quality. Current tool retrieval methods primarily focus on refining the ranking list of tools and directly packaging a fixed number of top-ranked tools as the tool set. However, these approaches often fail to equip LLMs with the optimal set of tools prior to execution, since the optimal number of tools for different tasks could be different, resulting in inefficiencies such as redundant or unsuitable tools, which impede immediate access to the most relevant tools. This paper addresses the challenge of recommending precise toolsets for LLMs. We introduce the problem of tool recommendation, define its scope, and propose a novel Precision-driven Tool Recommendation (PTR) approach. PTR captures an initial, concise set of tools by leveraging historical tool bundle usage and dynamically adjusts the tool set by performing tool matching, culminating in a multi-view-based tool addition. Additionally, we present a new dataset, RecTools, and a metric, TRACC, designed to evaluate the effectiveness of tool recommendation for LLMs. We further validate our design choices through comprehensive experiments, demonstrating promising accuracy across two open benchmarks and our RecTools dataset.
This work explores expanding the capabilities of large language models (LLMs) pretrained on text to generate 3D meshes within a unified model. This offers key advantages of (1) leveraging spatial knowledge already embedded in LLMs, derived from textual sources like 3D tutorials, and (2) enabling conversational 3D generation and mesh understanding. A primary challenge is effectively tokenizing 3D mesh data into discrete tokens that LLMs can process seamlessly. To address this, we introduce LLaMA-Mesh, a novel approach that represents the vertex coordinates and face definitions of 3D meshes as plain text, allowing direct integration with LLMs without expanding the vocabulary. We construct a supervised fine-tuning (SFT) dataset enabling pretrained LLMs to (1) generate 3D meshes from text prompts, (2) produce interleaved text and 3D mesh outputs as required, and (3) understand and interpret 3D meshes. Our work is the first to demonstrate that LLMs can be fine-tuned to acquire complex spatial knowledge for 3D mesh generation in a text-based format, effectively unifying the 3D and text modalities. LLaMA-Mesh achieves mesh generation quality on par with models trained from scratch while maintaining strong text generation performance.
Recent advancements in large language models(LLMs), such as GPT-4 and GPT-4o, have shown exceptional performance, especially in languages with abundant resources like English, thanks to extensive datasets that ensure robust training. Conversely, these models exhibit limitations when processing under-resourced languages such as Chinese and Korean, where issues including hallucinatory responses remain prevalent. This paper traces the roots of these disparities to the tokenization process inherent to these models. Specifically, it explores how the tokenizers vocabulary, often used to speed up the tokenization process and reduce tokens but constructed independently of the actual model training data, inadequately represents non-English languages. This misrepresentation results in the propagation of under-trained or untrained tokens, which perpetuate biases and pose serious concerns related to data security and ethical standards. We aim to dissect the tokenization mechanics of GPT-4o, illustrating how its simplified token-handling methods amplify these risks and offer strategic solutions to mitigate associated security and ethical issues. Through this study, we emphasize the critical need to rethink tokenization frameworks to foster more equitable and secure AI technologies. The code and data are available at //github.com/yeyimilk/LLMGPT4o
Key-Value (KV) caching is a common technique to enhance the computational efficiency of Large Language Models (LLMs), but its memory overhead grows rapidly with input length. Prior work has shown that not all tokens are equally important for text generation, proposing layer-level KV cache compression to selectively retain key information. Recognizing the distinct roles of attention heads in generation, we propose HeadKV, a head-level KV cache compression method, and HeadKV-R2, which leverages a novel contextual reasoning ability estimation for compression. Our approach operates at the level of individual heads, estimating their importance for contextual QA tasks that require both retrieval and reasoning capabilities. Extensive experiments across diverse benchmarks (LongBench, LooGLE), model architectures (e.g., Llama-3-8B-Instruct, Mistral-7B-Instruct), and long-context abilities tests demonstrate that our head-level KV cache compression significantly outperforms strong baselines, particularly in low-resource settings (KV size = 64 & 128). Notably, our method retains just 1.5% of the KV cache while achieving 97% of the performance of the full KV cache on the contextual question answering benchmark.Codes are available at //github.com/FYYFU/HeadKV
Deep long-tailed recognition has been widely studied to address the issue of imbalanced data distributions in real-world scenarios. However, there has been insufficient focus on the design of neural architectures, despite empirical evidence suggesting that architecture can significantly impact performance. In this paper, we attempt to mitigate long-tailed issues through architectural improvements. To simplify the design process, we utilize Differential Architecture Search (DARTS) to achieve this goal. Unfortunately, existing DARTS methods struggle to perform well in long-tailed scenarios. To tackle this challenge, we introduce Long-Tailed Differential Architecture Search (LT-DARTS). Specifically, we conduct extensive experiments to explore architectural components that demonstrate better performance on long-tailed data and propose a new search space based on our observations. This ensures that the architecture obtained through our search process incorporates superior components. Additionally, we propose replacing the learnable linear classifier with an Equiangular Tight Frame (ETF) classifier to further enhance our method. This classifier effectively alleviates the biased search process and prevents performance collapse. Extensive experimental evaluations demonstrate that our approach consistently improves upon existing methods from an orthogonal perspective and achieves state-of-the-art results with simple enhancements.
We present the R package MSTest, which implements hypothesis testing procedures to identify the number of regimes in Markov switching models. These models have wide-ranging applications in economics, finance, and numerous other fields. The MSTest package includes the Monte Carlo likelihood ratio test procedures proposed by Rodriguez-Rondon and Dufour (2024), the moment-based tests of Dufour and Luger (2017), the parameter stability tests of Carrasco, Hu, and Ploberger (2014), and the likelihood ratio test of Hansen (1992). Additionally, the package enables users to simulate and estimate univariate and multivariate Markov switching and hidden Markov processes, using the expectation-maximization (EM) algorithm or maximum likelihood estimation (MLE). We demonstrate the functionality of the MSTest package through both simulation experiments and an application to U.S. GNP growth data.
Recent advancements in Large Language Models (LLMs) have expanded their capabilities to multimodal contexts, including comprehensive video understanding. However, processing extensive videos such as 24-hour CCTV footage or full-length films presents significant challenges due to the vast data and processing demands. Traditional methods, like extracting key frames or converting frames to text, often result in substantial information loss. To address these shortcomings, we develop OmAgent, efficiently stores and retrieves relevant video frames for specific queries, preserving the detailed content of videos. Additionally, it features an Divide-and-Conquer Loop capable of autonomous reasoning, dynamically invoking APIs and tools to enhance query processing and accuracy. This approach ensures robust video understanding, significantly reducing information loss. Experimental results affirm OmAgent's efficacy in handling various types of videos and complex tasks. Moreover, we have endowed it with greater autonomy and a robust tool-calling system, enabling it to accomplish even more intricate tasks.
Transformer-based models have achieved remarkable success in various Natural Language Processing (NLP) tasks, yet their ability to handle long documents is constrained by computational limitations. Traditional approaches, such as truncating inputs, sparse self-attention, and chunking, attempt to mitigate these issues, but they often lead to information loss and hinder the model's ability to capture long-range dependencies. In this paper, we introduce ChuLo, a novel chunk representation method for long document classification that addresses these limitations. Our ChuLo groups input tokens using unsupervised keyphrase extraction, emphasizing semantically important keyphrase based chunk to retain core document content while reducing input length. This approach minimizes information loss and improves the efficiency of Transformer-based models. Preserving all tokens in long document understanding, especially token classification tasks, is especially important to ensure that fine-grained annotations, which depend on the entire sequence context, are not lost. We evaluate our method on multiple long document classification tasks and long document token classification tasks, demonstrating its effectiveness through comprehensive qualitative and quantitative analyses.
This article presents the affordances that Generative Artificial Intelligence can have in disinformation context, one of the major threats to our digitalized society. We present a research framework to generate customized agent-based social networks for disinformation simulations that would enable understanding and evaluation of the phenomena whilst discussing open challenges.
The problem of Multiple Object Tracking (MOT) consists in following the trajectory of different objects in a sequence, usually a video. In recent years, with the rise of Deep Learning, the algorithms that provide a solution to this problem have benefited from the representational power of deep models. This paper provides a comprehensive survey on works that employ Deep Learning models to solve the task of MOT on single-camera videos. Four main steps in MOT algorithms are identified, and an in-depth review of how Deep Learning was employed in each one of these stages is presented. A complete experimental comparison of the presented works on the three MOTChallenge datasets is also provided, identifying a number of similarities among the top-performing methods and presenting some possible future research directions.
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191