Skip to content

Watch-Later/paper-reading-2

 
 

Repository files navigation

Paper Reading -- Deep Learning Infra

比做算法的懂工程落地,比做工程的懂算法模型。

  1. 编程: c++ / CUDA / 汇编 / python / Shell
  2. 算法: deep learning / CV / NLP etc,训练框架,推理部署
  3. 工程: 硬件体系结构,OS & linux kernel, 分布式 & k8s 集群,存储
  4. 性能优化: 并行优化,deep learning 优化,profile 工具

X 的 100 问

Awesome Online Tools

table url
From #online-tools

AI Infra 设计

Deep Learning

Tutorials

Papers

计算机体系结构

CPU & Memory

全都用大量篇幅讲存储器的层次结构。

我的笔记:

GPU & CUDA 编程

Reading List

我的笔记

Nvidia Docs

汇编

  • x86 汇编
  • MIPS 汇编

编译原理

  1. Compilers Notes Map

编程语言

C++ 开发

Reading List

我的笔记地图 Cpp Notes Map

Python & Shell

性能优化

TODO 补充 deep learning 的高性能计算

架构设计

自动驾驶

Tutorials

自动驾驶 Lectures

Paper - Motion Planning 决策规划

Paper - Control 控制系统

可视化

NLP

Basic Models

  • Speech and Language Processing | pdf
  • A Stochastic Parts Program and Noun Phrase Parser for Unrestricted Text | pdf
  • A Statistical Approach to Machine Translation | pdf
  • Spam Filtering with Naive Bayes – Which Naive Bayes? | pdf

Undirected Graphical Model

alias: Markov network, Markov Random Field (MRF)

HMM, MEMM, CRF

  • introduction to Hidden Markov Model pdf Chap.8 Part-of-Speech Tagging
  • Conditional Random Fields - Probabilistic Models for Segmenting and Labeling Sequence Data | pdf

Near-duplicate Detection

  • N-gram Similarity and Distance | pdf
  • A Closer Look at Skip-gram Modelling | pdf
  • Similarity Estimation Techniques from Rounding Algorithms | pdf
  • Detecting Near-Duplicates for Web Crawling | pdf

搜索引擎

  • Web search for a planet: The Google cluster architecture | pdf

图理论

  • Finding and evaluating community structure in networks | pdf
  • Fast algorithm for detecting community structure in networks | pdf

数据库系统

papers

  • Bitcask - A Log-Structured Hash Table for Fast Key/Value Data | pdf | notes
  • The Google File System | pdf | notes
  • Bigtable: A Distributed Storage System for Structured Data | pdf
  • The Chubby lock service for loosely-coupled distributed systems | pdf

分布式系统

https://github.com/misc-codes/software-architecture-books

papers

  • Brewer’s Conjecture and the Feasibility of Consistent, Available, Partition-Tolerant Web | pdf | notes
  • BASE: An Acid Alternative | pdf | notes
  • Eventually consistent | pdf
  • Conflict-free Replicated Data Types | pdf
  • The Byzantine Generals Problem | pdf
  • The Part-Time Parliament | pdf
  • Paxos Made Simple | pdf
  • MapReduce: Simplified Data Processing on Large Clusters | pdf | notes

开发环境

git/vim/tmux/docker configs

其他

About

比做算法的懂工程落地,比做工程的懂算法模型。

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Jupyter Notebook 99.8%
  • Other 0.2%