预训练语言模型(PLMs)是在大规模语料库上以自监督方式进行预训练的语言模型。在过去的几年中，这些PLM从根本上改变了自然语言处理社区。在本教程中，我们旨在从两个角度提供广泛而全面的介绍:为什么这些PLM有效，以及如何在NLP任务中使用它们。本教程的第一部分对PLM进行了一些有见地的分析，部分解释了PLM出色的下游性能。第二部分首先关注如何将对比学习应用于PLM，以改进由PLM提取的表示，然后说明如何在不同情况下将这些PLM应用于下游任务。这些情况包括在数据稀缺的情况下对PLM进行微调，以及使用具有参数效率的PLM。我们相信，不同背景的与会者会发现本教程内容丰富和有用。

DITA 1.3说明文件，描述出现三个版本的DITA 1.3标准的原因，是为了不同类型的受众，交付了基础版本，技术版本和培训版本。

Database Engines on Multicores, Why Parallelize When You Can Distribute.pdf Multicore computers pose a substantial challenge to infrastructure software such as operating systems or databases. Such software typically evolves slower than the underlying hardware, and with multicore it faces structural limitations that can be solved only with radical architectural changes. In this paper we argue that, as has been suggested for operating systems, databases could treat multicore architectures

数字信号处理教学课件：Chapter 0 Why DSP.ppt

5 Why 分析法培训资料及相关的报告模板

没有广告水印，自编目录书签方便阅读

TeX's author, a famous professor, promoted Literate programming. This SHORT article explain WHY? It's only 14 pages, it change the view of a programmer. Literate programming is a style in which the design of the code refects that the human reader is as important as the machine reader.

Making Software - What Really Works, and Why We Believe It. Does the MMR vaccine cause autism? Does watching violence on TV make children more violent? Are some programming languages better than others? People argue about these questions every day. Every serious attempt to answer the first two questions relies on the scientific method: careful collection of evidence, and impartial evaluation of its implications. Until recently, though, only a few people have tried to apply these techniques to the third. When it comes to computing, it often seems that a couple glasses of beer and an anecdote about a startup in Warsaw are all the “evidence” most programmers expect. That is changing, thanks in part to the work of the contributors to this book. Drawing on fields as diverse as data mining, cognitive psychology, and sociology, they and their colleagues are creating an evidence-based approach to software engineering. By gathering evidence drawn from a myriad of primary sources and analyzing the results, they are shedding new light onto some vexing questions of software development. What do most programmers get wrong in their first job? Does test-driven development lead to better code? What about pair programming, or code reviews? Is it possible to predict the likely number of bugs in a piece of code before it’s released? If so, how?

#1.1_Why__(PyTorch_tutorial_神经网络_教学)

#1.1_Why_Linux__(Linux_机器学习_教程教学_tutorial)