paper-with-me

홈 › Papers

LongLLaVA: Scaling Multi-modal LLMs to 1000 Images Efficiently via a Hybrid Architecture

2024-09-04 · Xidong Wang, Dingjie Song, Shunian Chen, Chen Zhang, Benyou Wang

Expanding the long-context capabilities of Multi-modal Large Language Models~(MLLMs) is crucial for video understanding, high-resolution image understanding, and multi-modal agents. This involves a series of systematic optimizations, including model architecture, data construction and training strategy, particularly addressing challenges such as \textit{degraded performance with more images} and \textit{high computational costs}. In this paper, we adapt the model architecture to a hybrid of Mamba and Transformer blocks, approach data construction with both temporal and spatial dependencies among multiple images and employ a progressive training strategy. The released model \textbf{LongLLaVA}~(\textbf{Long}-Context \textbf{L}arge \textbf{L}anguage \textbf{a}nd \textbf{V}ision \textbf{A}ssistant) is the first hybrid MLLM, which achieved a better balance between efficiency and effectiveness. LongLLaVA not only achieves competitive results across various benchmarks, but also maintains high throughput and low memory consumption. Especially, it could process nearly a thousand images on a single A100 80GB GPU, showing promising application prospects for a wide range of tasks.

📄 PDF Abstract BibTeX arXiv:2409.02889

Code (1)

freedomintelligence/longllava 공식 구현 pytorch

Tasks

GPUMambaVideo Understanding

Methods 이 논문이 사용한 방법론

Attention 설명 없음
BPE Byte Pair Encoding, or BPE, is a subword segmentation algorithm that encodes rare and unknown words as sequences of subword units. The intuition is that various word…
Absolute Position Encodings Absolute Position Encodings are a type of position embeddings for [Transformer-based models] where positional encodings are…
Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…
Mamba Foundation models, now powering most of the exciting applications in deep learning, are almost universally based on the Transformer architecture and its core attention module.…
Label Smoothing Label Smoothing is a regularization technique that introduces noise for the labels. This accounts for the fact that datasets may have mistakes in them, so maximizing the…
Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
Adam 설명 없음

Similar Papers 제목 키워드 기반

NaViL: Rethinking Scaling Properties of Native Multimodal Large Language Models under Data Constraints

2025-10-09 · Changyao Tian, Hao Li, Gen Luo, Xizhou Zhu 외 arxiv

Compositional training has been the de-facto paradigm in existing Multimodal Large Language Models (MLLMs), where pre-trained vision encoders are connected with pre-trained LLMs through continuous multimodal pre-training…

Plug-and-Play Grounding of Reasoning in Multimodal Large Language Models

2024-03-28 · Jiaxing Chen, Yuxuan Liu, Dehu Li, Xiang An 외

The rise of Multimodal Large Language Models (MLLMs), renowned for their advanced instruction-following and reasoning capabilities, has significantly propelled the field of visual reasoning. However, due to limitations i…

Instruction FollowingVisual Reasoning

Can MLLMs Reason in Multimodality? EMMA: An Enhanced MultiModal ReAsoning Benchmark

2025-01-09 · Yunzhuo Hao, Jiawei Gu, Huichen Will Wang, Linjie Li 외

The ability to organically reason over and with both text and images is a pillar of human intelligence, yet the ability of Multimodal Large Language Models (MLLMs) to perform such multimodal reasoning remains under-explo…

Multimodal Reasoning

Test-time Scaling over Perception: Resolving the Grounding Paradox in Thinking with Images

2026-04-13 · Zheng Jiang, Yiming Chen, Nan He, Jiahui Chen 외 arxiv

Recent multimodal large language models (MLLMs) have begun to support Thinking with Images by invoking visual tools such as zooming and cropping during inference. Yet these systems remain brittle in fine-grained visual r…

Multimodal ReasoningVisual Reasoning

GME: Improving Universal Multimodal Retrieval by Multimodal LLMs

2024-12-22 · Xin Zhang, Yanzhao Zhang, Wen Xie, Mingxin Li 외

Universal Multimodal Retrieval (UMR) aims to enable search across various modalities using a unified model, where queries and candidates can consist of pure text, images, or a combination of both. Previous work has attem…

Retrieval