526互联

gpu-accelerated accelerated efficient subgraph

SHARPNESS-AWARE MINIMIZATION FOR EFFICIENTLY IMPROVING GENERALIZATION论文阅读笔记

Intro 在训练集上最小化损失很可能导致泛化性低，因为当今模型的过参数化会导致training loss的landscape异常复杂且非凸，包含很多local/global minima，因此优化器的选择至关重要。loss landscape的几何性质（特别是minima的flatness）与泛化 ......
SHARPNESS-AWARE GENERALIZATION MINIMIZATION EFFICIENTLY SHARPNESS

AMOS: Enabling Automatic Mapping for Tensor Computations On Spatial Accelerators with Hardware Abstraction

AMOS: Enabling Automatic Mapping for Tensor Computations On Spatial Accelerators with Hardware Abstraction Abstract 为了实现性能提升，硬件专用化是一个趋势。空间硬件加速器利用专门的层次 ......
Accelerators Computations Abstraction Automatic Enabling

LocPatcH An efficient long-read hybrid error correction algorithm based on local pHMM

该文档主要介绍了一种基于装配的方法和概率隐藏马尔科夫模型 (pHMM) 用于纠正长读序列的错误。文档详细描述了对酵母数据进行实验的结果、纠正方法的拓扑结构以及实验设置和数据集。 这种基于装配的纠正方法相对于直接纠正存在哪些优势？ pHMM 的拓扑结构是怎样的？ 在实验中使用了什么样的数据集？ 提示： ......
correction efficient algorithm long-read LocPatcH

UNetFormer: A UNet-like transformer for efficient semantic segmentation of remote sensing urban scene imagery

UNetFormer: A UNet-like transformer for efficient semantic segmentation of remote sensing urban scene imagery * Authors: [[Libo Wang]], [[Rui Li]], [[ ......
segmentation transformer UNetFormer UNet-like efficient

Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network

Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network * Authors: [[Wenzhe Shi]], [[Jose Caballer ......
Super-Resolution Convolutional Resolution Real-Time Efficient

High-Efficiency Lossy Image Coding Through Adaptive Neighborhood Information Aggregation

目录简介创新点内容Entropy Coding Using Multistage Context Model模型结构残差邻域注意力块Residual Neighborhood Attention Block RNAB激活函数 高斯误差线性单元激活函数GELU并行解码 简介 创新点 Integrate ......
High-Efficiency Neighborhood Aggregation Information Efficiency

【LLMOps】Accelerate & DeepSpeed多卡使用

介绍 目前大模型微调主要方案是 LLaMA-Factory LLaMA-Factory中，提供了两种多卡框架：Accelerate、DeepSpeed Accelerate 依赖 accelerate==0.24.1transformers==4.34.1datasets==2.14.7tiktok ......
Accelerate DeepSpeed LLMOps amp

使用Accelerate库在多GPU上进行LLM推理

大型语言模型(llm)已经彻底改变了自然语言处理领域。随着这些模型在规模和复杂性上的增长，推理的计算需求也显著增加。为了应对这一挑战利用多个gpu变得至关重要。 所以本文将在多个gpu上并行执行推理，主要包括：Accelerate库介绍，简单的方法与工作代码示例和使用多个gpu的性能基准测试。 本文 ......
Accelerate GPU LLM

【HFSS】仿真时radiation efficiency（辐射效率）大于1

![image](https://img2023.cnblogs.com/blog/2603751/202311/2603751-20231121124817927-404911195.png) ![image](https://img2023.cnblogs.com/blog/2603751/20... ......
efficiency radiation 效率 HFSS

TALLRec: An Effective and Efficient Tuning Framework to Align Large Language Model with Recommendation

目录概TallRec代码 Bao K., Zhang J., Zhang Y., Wang W., Feng F. and He X. TALLRec: An effective and efficient tuning framework to align large language model ......
Recommendation Effective Efficient Framework Language

2023CVPR_Efficient Frequency Domain-based Transformers for High-Quality Image Deblurring

一. Motivation 1. Transformer在解决全局表现很好，但是复杂度很高，主要体现在QK的乘积: (We note that the scaled dot-product attention computation is actually to estimate the corre ......
CVPR_Efficient Domain-based Transformers High-Quality Deblurring

Checkerboard Context Model for Efficient Learned Image Compression

目录AbstractIntroductionPreliminary 初步介绍Variational Image Compression with Hyperprior(超先验变分图像压缩)Autoregressive Context(自回归上下文模型)Parallel Context Modelin ......
Checkerboard Compression Efficient Context Learned

ELIC: Efficient Learned Image Compression with Unevenly Grouped Space-Channel Contextual Adaptive Coding

abstruct \(\quad\) 受能量压缩表现的启发，提出了不均匀通道情况自适应编码.结合不均匀分组模型和现有上下文模型，获得一种空间通道上下文自适应模型，来提高编码性能，而不影响其运行时间。 \(\quad\)这种模型支持预览解码和渐进解码。 introduction 学习图像压缩中最重要的 ......
Space-Channel Compression Contextual Efficient Adaptive

【CVPR2023】Efficient and Explicit Modelling of Image Hierarchies for Image Restoration

> 论文：https://readpaper.com/paper/4728855966703960065 代码：https://github.com/ofsoundof/GRL-Image-Restoration 这个论文的代码地址叫GRL，意思是 Global, Regional, Local 的 ......
Image Hierarchies Restoration Efficient Modelling

2023ACMMM_Mutual Information-driven Triple Interaction Network for Efficient Image Dehazing

一. Motivation 之前网络存在的缺点： 1. 使用的有限的频域信息 2. 不充足的信息交互 ： (1) 第一阶段的输出直接作为第二阶段的输入，忽略了中间特征从早期到后期的传播 (2) 在编码器解码器结构同尺度之间进行特征融合，忽略了阶段内和跨阶段的跨尺度信息交换 3. 严重的特征冗余：中间 ......
Information-driven ACMMM_Mutual Information Interaction Efficient

Metropolis Algorithms for Representative Subgraph Sampling

目录概主要内容Metropolis graph sampling H\¨{u}bler C. and Kriegel H., Borgwardt K. and Ghahramani Z. Metropolis algorithms for representative subgraph sampli ......
Representative Metropolis Algorithms Subgraph Sampling

论文阅读：A Lightweight Knowledge Graph Embedding Framework for Efficient Inference and Storage

ABSTRACT 现存的KGE方法无法适用于大规模的图（由于存储和推理效率的限制） 作者提出了一种LightKG框架： 自动的推断出码本codebooks和码字codewords，为每个实体生成合适的embedding。 同时，框架中包含残差模块来实现码本的多样性，并且包含连续函数来近似的实现码字的 ......
Lightweight Knowledge Embedding Efficient Framework

Efficient Off-Policy Meta-Reinforcement Learning via Probabilistic Context Variables

郑重声明：原文参见标题，如有侵权，请联系作者，将会撤销发布！ Proceedings of the 36th International Conference on Machine Learning, PMLR 97:5331-5340, 2019 ......
Meta-Reinforcement Reinforcement Probabilistic Off-Policy Efficient

[VLDB 2012]Efficient Subgraph Matching on Billion Node Graphs

[VLDB 2012]Efficient Subgraph Matching on Billion Node Graphs 重点了解实现star-join的具体过程。 分解query和STwigs排序 文中把star叫做STwigs，每一个STwigs查询为\(q=(r, L)\)，其中r是跟节点标 ......
Efficient Subgraph Matching Billion Graphs

Proj CDeepFuzz Paper Reading: SparseProp: Efficient Sparse Backpropagation for Faster Training of Neural Networks

## Abstract 本文：SparseProp Github: https://github.com/IST-DASLab/sparseprop Task: a back-propagation algo for sparse training data, a fast vectorized i ......
Backpropagation SparseProp CDeepFuzz Efficient Networks

​MPDIoU: A Loss for Efficient and Accurate Bounding Box Regression

​MPDIoU: A Loss for Efficient and Accurate Bounding Box Regression MPDIoU:一个有效和准确的边界框损失回归函数 摘要 边界框回归(Bounding box regression, BBR)广泛应用于目标检测和实例分割，是目标定位 ......
Regression Efficient Accurate Bounding MPDIoU

[VLDBJ 2019]Distributed Subgraph Matching on Timely Dataflow

# [VLDBJ 2019]Distributed Subgraph Matching on Timely Dataflow **只关注这篇中的subgraph matching的内容** ## 定义 $g = (V_g, E_g, L_g)$分别表示点、边，以及把任意点或边映射成label的函数。 ......
Distributed Dataflow Subgraph Matching Timely

大模型训练框架deepspeed和accelerate

引子 DeepSpeed是由Microsoft提供的分布式训练工具，旨在支持更大规模的模型和提供更多的优化策略和工具。与其他框架相比，DeepSpeed支持更大规模的模型和提供更多的优化策略和工具。其中，主要优势在于支持更大规模的模型、提供了更多的优化策略和工具（例如 ZeRO 和 Offload ......
accelerate deepspeed 框架 模型

Proj CDeepFuzz Paper Reading: Aries: Efficient Testing of Deep Neural Networks via Labeling-Free Accuracy Estimation

## Abstract 背景： 1. the de facto standard to assess the quality of DNNs in the industry is to check their performance (accuracy) on a collected set of ......
Labeling-Free Estimation CDeepFuzz Efficient Accuracy

[SIGIR 2023] Subgraph Search over Neural-Symbolic Graphs

# [SIGIR 2023] Subgraph Search over Neural-Symbolic Graphs ## 总结 ## 研究的问题 在包含非结构化数据（图像、视频、文本等）的神经符号数据库（neural-symbolic graph datasets）上如何进行高效的神经符号子图匹配 ......
Neural-Symbolic Subgraph Symbolic Neural Graphs

[论文精读][计算生物][蛋白质预训练表示]Data-Efficient Protein 3D Geometric Pretraining via Refinement of Diffused Protein Structure Decoy

笔者正在调研市面上的蛋白表示方法，论文方法过于数理的部分会被抽象带过。 ## Basic Information: * Title: Data-Efficient Protein 3D Geometric Pretraining via Refinement of Diffused Protein St ......
Protein Data-Efficient 蛋白质 Pretraining Refinement

[SIGMOD 2020]In-Memory Subgraph Matching An In-depth Study

# In-Memory Subgraph Matching: An In-depth Study 一篇subgraph matching的survey ## 总结 ![img](https://img2023.cnblogs.com/blog/2988279/202308/2988279-20230 ......
In-Memory Matching In-depth Subgraph SIGMOD

NEURAL SUBGRAPH MATCHING

# NEURAL SUBGRAPH MATCHING ## 总结 ## 问题定义 给定一个查询图，判断该图是不是一个大图的子图。如果图中有点和边的特征，就要都匹配上。 ## 动机 同构问题是NP完全的，已有的方法虽然能匹配很大的target图，但query图会很小。 ## 模型框架 分为embedd ......
MATCHING SUBGRAPH NEURAL

Efficient and Accurate Diagnostic Tool

Diagnostic tools play a crucial role in the automotive industry, allowing technicians to accurately identify and troubleshoot vehicle issues. Among th ......
Diagnostic Efficient Accurate Tool and

Classical Management: emphasized rationality and making organizations and workers as efficient as possible

Classical approach: First studies of management, which emphasized: * rationality * making organizations and workers as efficient as possible **Max Web ......
organizations rationality Management emphasized Classical
共61篇  :1/3页 首页上一页1下一页尾页