Tags

keep hungry keep foolish
生活

首篇博客

First blog

docker

Kubernetes入门实践

Kubernetes introduction

Linux信号机制与docker应用

Linux signal in docker

k8s

Kubernetes入门实践

Kubernetes introduction

Linux信号机制与docker应用

Linux signal in docker

linux

Linux信号机制与docker应用

Linux signal in docker

react

React router 入门

Introduction to react router

React hooks 入门

React hooks

React生命周期介绍

React lifecycle

React与Knockout共存探索

React and knockout

knockout

React与Knockout共存探索

React and knockout

frontend

虚拟 DOM 与 diff 算法

Virtual DOM and diff algorithm

PWA 与 Github Page

PWA and Github page

Form 的优化与思考

Form optimization and thinking

Canvas 基础应用

Introduction to canvas

React router 入门

Introduction to react router

Airbnb React/JSX 编码规范

Airbnb React/JSX Style Guide

This 关键字

Keyword this

ES6 概述

ES6

Flex 布局

Flex

React hooks 入门

React hooks

Less与Sass使用

less and sass

Promise 异步调用

Promise in javascript

React生命周期介绍

React lifecycle

React与Knockout共存探索

React and knockout

python

深入源码，洞察迭代 8 年的 html 文本转换库

Dive into the source code and gain insights into the html text conversion library that has been iterating for 8 years

深入源码，洞察迭代 14 年的高质量 html 提取方案

Go deep into the source code and gain insight into the high-quality HTML extraction solution that has been iterated for 14 years

Dify框架增强：知识库检索能力提升探索与实践

Enhancement of Dify framework: exploration and practice of improving the retrieval ability of knowledge base

深入 Dify 源码，洞察 Dify RAG 切片机制实现细节

Go deep into the Dify source code and gain insight into the implementation details of the Dify RAG slicing mechanism

来自工业界的开源知识库 RAG 项目结构化文件解析方案比较

Comparison of structured file parsing solutions from open source RAG projects in the industry

深入 Dify 源码，洞察 Dify RAG 默认机制

Dive into the Dify source code and gain insight into the Dify RAG default mechanism

深入 Dify 源码，定位知识库检索的大模型调用异常

Go deep into the Dify source code and locate large model call anomalies in knowledge base retrieval

基于开源项目二次开发可行方案

The solution for secondary development based on open source projects

Dify 与 Xinference 最佳组合 GPU 环境部署全流程

The whole process of deploying the combination of Dify and Xinference in GPU environment

来自工业界的开源知识库 RAG 项目最全细节对比

The most complete comparison of open source RAG projects from the industry

官方技术文档查询大模型：快速搭建与实践记录

Official technical document query large model: rapid construction and practice records

一款纯 js 实现的大模型应用服务 FastGPT 解读

Interpretation of FastGPT, a large model application service implemented in pure js

Github 上 Star 数最多的大模型应用基础服务 Dify 深度解读（一）

Dify source code analysis of the large model application basic service with the largest number of stars on Github (1)

来自工业界的知识库 RAG 服务(五)，模块化知识库 GoMate 实现方案详解

RAG services from industry (5), detailed explanation of modular knowledge base GoMate implementation plan

从 0 打造私有知识库 RAG Benchmark 完整实践

The practical process of building a private RAG Benchmark from scratch

来自工业界的知识库 RAG 服务(四)，FinGLM 竞赛冠军项目详解

Knowledge base RAG service from the industry (4), detailed explanation of the FinGLM competition championship project

来自工业界的知识库 RAG 服务(三)，FinGLM 竞赛获奖项目详解

Knowledge base RAG service from the industry (3), detailed explanation of the winning projects of the FinGLM competition

来自学术界的知识库 RAG 调优方案实践（一）

RAG tuning solution practice from academia (1)

从《红楼梦》的视角看大模型知识库 RAG 服务的 Rerank 调优

Rerank tuning of RAG service in large model knowledge base from the perspective of 'Dream of Red Mansions'

来自工业界的知识库 RAG 服务(二)，RagFlow 源码全流程深度解析

Knowledge base RAG service from the industry (2), in-depth analysis of the entire RagFlow source code process

来自工业界的 RAG 服务，有道 QAnything 源码全流程深度解析

RAG service from the industry, Youdao QAnything source code in-depth analysis of the entire process

一种与众不同的 RAG 架构探索

An exploration of a distinctive RAG architecture

向量数据库 Faiss 的实践与增强探索

Practice and enhanced exploration of vector database Faiss

从开发到部署，搭建离线私有大模型知识库

From development to deployment, build an offline private large model knowledge base

人 GPT 协作，从 0 手撸一个量化交易回测平台

Collaborate with GPT，create a quantitative trading backtesting platform from scratch

难以训练的深度学习

Difficult to train deep learning

从 0 手撸一个 pytorch

Implementing a simple version of pytorch from 0

FATE 作业执行全流程追踪

Full process tracking of FATE job execution

最简安全纵向联邦学习实现探索

Exploring the most simplified implementation of secure vertical federated learning

深入探索 FATE 纵向联邦学习安全方案

In-depth exploration of FATE vertical federated learning security solution

深入探索 FATE 纵向联邦学习模型设计方案

In-depth exploration of the FATE vertical federated learning model design solution

FATE 纵向联邦学习实现探索

FATE vertical federated learning implementation exploration

深入探索联邦学习框架 Flower

A deep dive into the federated learning framework Flower

联邦学习 Non-IID 算法实现

Federated Learning Non-IID Algorithm implementation

GPT Engineer 实践与源码解析

GPT engineer practice and source code analysis

medical-labelme 升级方案概述

Overview of the medical-labelme upgrade plan

自研医疗图像标注工具 medical-labelme

Self-developed medical image annotation tool medical-labelme

DICOM 图像探索

DICOM image exploration

FATE Eggroll 源码解析

FATE Eggroll source code analysis

联邦学习下线性回归算法实现概述

Overview of the implementation of linear regression algorithm under federated learning

FATE Board 执行流程探索

FATE Board Execution Process Exploration

基于 Pre-commit 的代码风格统一实践

Unified Code Style Practice Based on Pre-commit

基于 OpenAPI 的 Web 页面自动渲染实践

Practice of Automatic Rendering of Web Pages Based on OpenAPI

FATE Flow 源码解析 - 日志输出机制

FATE Flow source code analysis - log mechanism

FATE Flow 源码解析 - 资源分配流程

FATE Flow source code analysis - resource allocation process

FATE Flow 源码解析 - 作业提交处理流程

FATE Flow source code analysis - task submission process

基于 Grpc interceptors 的请求跟踪探索

Exploration of request tracking based on grpc interceptors

基于 cookiecutter 的 python 项目模板

Python project template based on cookiecutter

Python 中的时区问题

Timezone problem in python

Poetry 包管理

Python package management by poetry

基于 FastAPI 与 MongoDB 的数据流梳理

Data flow base on FastApi and MongoDB

生成器与协程

Generator and Coroutine

Requests cookie 源码分析

Requests cookie source codes reading

基于 Tushare 进行数据分析

Data analysis base on tushare

Pandas 数据分析

Data analysis with pandas

Requests 源码阅读

Requests source codes reading

RBAC 权限管理介绍与实现

Introduction and implementation of RBAC authority management

升级 Python3

upgrade to python3

flask 信号机制实现

Signal in flask

深入了解 gevent

Dive into gevent

Celery 3.x 升级至 celery 4.x

Celery 3.x upgrade to celery 4.x

从 Flask-script 迁移到 Click

Migrating from flask-script to click

PEP8 规范

PEP 8

Mock与模块引入问题

mock and module import

Celery框架介绍

dive into celery

单元测试mock模块介绍

mock

实用的装饰器库decorator

decorator

基于属性的测试库-hypothesis

hypothesis

python中的循环引用

circular import in python

自动化代码质量控制

automating code quality

logging入门

logging in python

Python虚拟环境管理

Virtual environment in python

Flask 0.1 源码解析

Dive into flask 0.1

Python上下文管理

Context managers in python

Local，localStack，localProxy深入解析

Local, localStack, localProxy in flask and werkzeug

Python优雅迭代器实现

Elegant iterator in python

Python字符编码介绍

Encoding in python

flask

Flask 0.1 源码解析

Dive into flask 0.1

Local，localStack，localProxy深入解析

Local, localStack, localProxy in flask and werkzeug

tools

git 大小写问题

Git ignorecase issue

Git笔记

dive into git

Oh my zsh 配置与插件

Configurations of oh my zsh

javascript

Promise 异步调用

Promise in javascript

skills

源码阅读

Source codes reading

代码整洁之道

Clean code

程序员的软技能

Soft skills

代码重构

Code refactor

books

代码整洁之道

Clean code

程序员的软技能

Soft skills

代码重构

Code refactor

css

Less与Sass使用

less and sass

deploy

Nginx rewrite 机制梳理

Nginx rewrite introduction

12 要素

12 factors

FastAPI

基于 FastAPI 与 MongoDB 的数据流梳理

Data flow base on FastApi and MongoDB

MongoDB

基于 FastAPI 与 MongoDB 的数据流梳理

Data flow base on FastApi and MongoDB

grpc

基于 Grpc interceptors 的请求跟踪探索

Exploration of request tracking based on grpc interceptors

grpc interceptors

基于 Grpc interceptors 的请求跟踪探索

Exploration of request tracking based on grpc interceptors

FATE

FATE Eggroll 源码解析

FATE Eggroll source code analysis

FATE Board 执行流程探索

FATE Board Execution Process Exploration

FATE Flow 源码解析 - 日志输出机制

FATE Flow source code analysis - log mechanism

FATE Flow 源码解析 - 资源分配流程

FATE Flow source code analysis - resource allocation process

FATE Flow 源码解析 - 作业提交处理流程

FATE Flow source code analysis - task submission process

Federated learning

FATE 作业执行全流程追踪

Full process tracking of FATE job execution

最简安全纵向联邦学习实现探索

Exploring the most simplified implementation of secure vertical federated learning

深入探索 FATE 纵向联邦学习安全方案

In-depth exploration of FATE vertical federated learning security solution

深入探索 FATE 纵向联邦学习模型设计方案

In-depth exploration of the FATE vertical federated learning model design solution

FATE 纵向联邦学习实现探索

FATE vertical federated learning implementation exploration

深入探索联邦学习框架 Flower

A deep dive into the federated learning framework Flower

联邦学习 Non-IID 算法实现

Federated Learning Non-IID Algorithm implementation

FATE Eggroll 源码解析

FATE Eggroll source code analysis

联邦学习下线性回归算法实现概述

Overview of the implementation of linear regression algorithm under federated learning

FATE Flow 源码解析 - 日志输出机制

FATE Flow source code analysis - log mechanism

FATE Flow 源码解析 - 资源分配流程

FATE Flow source code analysis - resource allocation process

FATE Flow 源码解析 - 作业提交处理流程

FATE Flow source code analysis - task submission process

JSON Schema

基于 OpenAPI 的 Web 页面自动渲染实践

Practice of Automatic Rendering of Web Pages Based on OpenAPI

OpenAPI

基于 OpenAPI 的 Web 页面自动渲染实践

Practice of Automatic Rendering of Web Pages Based on OpenAPI

Pre-commit

基于 Pre-commit 的代码风格统一实践

Unified Code Style Practice Based on Pre-commit

Machine learning

人 GPT 协作，从 0 手撸一个量化交易回测平台

Collaborate with GPT，create a quantitative trading backtesting platform from scratch

难以训练的深度学习

Difficult to train deep learning

从 0 手撸一个 pytorch

Implementing a simple version of pytorch from 0

FATE 作业执行全流程追踪

Full process tracking of FATE job execution

最简安全纵向联邦学习实现探索

Exploring the most simplified implementation of secure vertical federated learning

深入探索 FATE 纵向联邦学习安全方案

In-depth exploration of FATE vertical federated learning security solution

深入探索 FATE 纵向联邦学习模型设计方案

In-depth exploration of the FATE vertical federated learning model design solution

FATE 纵向联邦学习实现探索

FATE vertical federated learning implementation exploration

深入探索联邦学习框架 Flower

A deep dive into the federated learning framework Flower

联邦学习 Non-IID 算法实现

Federated Learning Non-IID Algorithm implementation

GPT Engineer 实践与源码解析

GPT engineer practice and source code analysis

联邦学习下线性回归算法实现概述

Overview of the implementation of linear regression algorithm under federated learning

Eggroll

FATE Eggroll 源码解析

FATE Eggroll source code analysis

RollSite

FATE Eggroll 源码解析

FATE Eggroll source code analysis

DICOM

medical-labelme 升级方案概述

Overview of the medical-labelme upgrade plan

自研医疗图像标注工具 medical-labelme

Self-developed medical image annotation tool medical-labelme

DICOM 图像探索

DICOM image exploration

medical-labelme

medical-labelme 升级方案概述

Overview of the medical-labelme upgrade plan

自研医疗图像标注工具 medical-labelme

Self-developed medical image annotation tool medical-labelme

aigc

GPT Engineer 实践与源码解析

GPT engineer practice and source code analysis

Quantitative trading

人 GPT 协作，从 0 手撸一个量化交易回测平台

Collaborate with GPT，create a quantitative trading backtesting platform from scratch

LLM

从开发到部署，搭建离线私有大模型知识库

From development to deployment, build an offline private large model knowledge base

人 GPT 协作，从 0 手撸一个量化交易回测平台

Collaborate with GPT，create a quantitative trading backtesting platform from scratch

langchain

来自工业界的知识库 RAG 服务(五)，模块化知识库 GoMate 实现方案详解

RAG services from industry (5), detailed explanation of modular knowledge base GoMate implementation plan

从 0 打造私有知识库 RAG Benchmark 完整实践

The practical process of building a private RAG Benchmark from scratch

来自工业界的知识库 RAG 服务(四)，FinGLM 竞赛冠军项目详解

Knowledge base RAG service from the industry (4), detailed explanation of the FinGLM competition championship project

来自工业界的知识库 RAG 服务(三)，FinGLM 竞赛获奖项目详解

Knowledge base RAG service from the industry (3), detailed explanation of the winning projects of the FinGLM competition

来自学术界的知识库 RAG 调优方案实践（一）

RAG tuning solution practice from academia (1)

向量数据库 Faiss 的实践与增强探索

Practice and enhanced exploration of vector database Faiss

从开发到部署，搭建离线私有大模型知识库

From development to deployment, build an offline private large model knowledge base

faiss

向量数据库 Faiss 的实践与增强探索

Practice and enhanced exploration of vector database Faiss

rag

来自工业界的知识库 RAG（六），独特的 RAG 框架 dsRAG 核心亮点解读

RAG services from industry (6)，an Interpretation of the Core Highlights of the RAG Framework dsRAG

深入源码，洞察迭代 8 年的 html 文本转换库

Dive into the source code and gain insights into the html text conversion library that has been iterating for 8 years

深入源码，洞察迭代 14 年的高质量 html 提取方案

Go deep into the source code and gain insight into the high-quality HTML extraction solution that has been iterated for 14 years

Dify框架增强：知识库检索能力提升探索与实践

Enhancement of Dify framework: exploration and practice of improving the retrieval ability of knowledge base

深入 Dify 源码，洞察 Dify RAG 切片机制实现细节

Go deep into the Dify source code and gain insight into the implementation details of the Dify RAG slicing mechanism

来自工业界的开源知识库 RAG 项目结构化文件解析方案比较

Comparison of structured file parsing solutions from open source RAG projects in the industry

深入 Dify 源码，洞察 Dify RAG 默认机制

Dive into the Dify source code and gain insight into the Dify RAG default mechanism

深入 Dify 源码，定位知识库检索的大模型调用异常

Go deep into the Dify source code and locate large model call anomalies in knowledge base retrieval

基于开源项目二次开发可行方案

The solution for secondary development based on open source projects

Dify 与 Xinference 最佳组合 GPU 环境部署全流程

The whole process of deploying the combination of Dify and Xinference in GPU environment

来自工业界的开源知识库 RAG 项目最全细节对比

The most complete comparison of open source RAG projects from the industry

官方技术文档查询大模型：快速搭建与实践记录

Official technical document query large model: rapid construction and practice records

一款纯 js 实现的大模型应用服务 FastGPT 解读

Interpretation of FastGPT, a large model application service implemented in pure js

Github 上 Star 数最多的大模型应用基础服务 Dify 深度解读（一）

Dify source code analysis of the large model application basic service with the largest number of stars on Github (1)

来自工业界的知识库 RAG 服务(五)，模块化知识库 GoMate 实现方案详解

RAG services from industry (5), detailed explanation of modular knowledge base GoMate implementation plan

从 0 打造私有知识库 RAG Benchmark 完整实践

The practical process of building a private RAG Benchmark from scratch

来自工业界的知识库 RAG 服务(四)，FinGLM 竞赛冠军项目详解

Knowledge base RAG service from the industry (4), detailed explanation of the FinGLM competition championship project

来自工业界的知识库 RAG 服务(三)，FinGLM 竞赛获奖项目详解

Knowledge base RAG service from the industry (3), detailed explanation of the winning projects of the FinGLM competition

来自学术界的知识库 RAG 调优方案实践（一）

RAG tuning solution practice from academia (1)

从《红楼梦》的视角看大模型知识库 RAG 服务的 Rerank 调优

Rerank tuning of RAG service in large model knowledge base from the perspective of 'Dream of Red Mansions'

来自工业界的知识库 RAG 服务(二)，RagFlow 源码全流程深度解析

Knowledge base RAG service from the industry (2), in-depth analysis of the entire RagFlow source code process

来自工业界的 RAG 服务，有道 QAnything 源码全流程深度解析

RAG service from the industry, Youdao QAnything source code in-depth analysis of the entire process

一种与众不同的 RAG 架构探索

An exploration of a distinctive RAG architecture

dspy

一种与众不同的 RAG 架构探索

An exploration of a distinctive RAG architecture

qanything

来自工业界的 RAG 服务，有道 QAnything 源码全流程深度解析

RAG service from the industry, Youdao QAnything source code in-depth analysis of the entire process

ragflow

来自工业界的知识库 RAG 服务(二)，RagFlow 源码全流程深度解析

Knowledge base RAG service from the industry (2), in-depth analysis of the entire RagFlow source code process

dify

No Code 的大模型应用跟踪与批量测试方案

No Code's large model application tracking and batch testing solution

你的大模型应用表现真的好吗？借助 Dify + Langfuse 一探究竟

Is your large model application really performing well? Find out with Dify + Langfuse

Dify框架增强：知识库检索能力提升探索与实践

Enhancement of Dify framework: exploration and practice of improving the retrieval ability of knowledge base

深入 Dify 源码，洞察 Dify RAG 切片机制实现细节

Go deep into the Dify source code and gain insight into the implementation details of the Dify RAG slicing mechanism

来自工业界的开源知识库 RAG 项目结构化文件解析方案比较

Comparison of structured file parsing solutions from open source RAG projects in the industry

深入 Dify 源码，洞察 Dify RAG 默认机制

Dive into the Dify source code and gain insight into the Dify RAG default mechanism

Github 上 Star 数最多的大模型应用基础服务 Dify 深度解读（一）

Dify source code analysis of the large model application basic service with the largest number of stars on Github (1)

FastGPT

一款纯 js 实现的大模型应用服务 FastGPT 解读

Interpretation of FastGPT, a large model application service implemented in pure js

agent

AIOps RAG 比赛获奖项目 EasyRAG 深度解读

In-depth interpretation of AIOps RAG competition winning project EasyRAG

RAG 分块长距离信息缺失，Late Chunking 值得试试

Long distance contextual information is missing in RAG, you can try Late Chunking

No Code 的大模型应用跟踪与批量测试方案

No Code's large model application tracking and batch testing solution

你的大模型应用表现真的好吗？借助 Dify + Langfuse 一探究竟

Is your large model application really performing well? Find out with Dify + Langfuse

给大模型加上记忆，深入探索 Mem0 项目

Adding memory to large models, exploring the Mem0 project

来自工业界的知识库 RAG（六），独特的 RAG 框架 dsRAG 核心亮点解读

RAG services from industry (6)，an Interpretation of the Core Highlights of the RAG Framework dsRAG

Dify框架增强：知识库检索能力提升探索与实践

Enhancement of Dify framework: exploration and practice of improving the retrieval ability of knowledge base

深入 Dify 源码，洞察 Dify RAG 切片机制实现细节

Go deep into the Dify source code and gain insight into the implementation details of the Dify RAG slicing mechanism

来自工业界的开源知识库 RAG 项目结构化文件解析方案比较

Comparison of structured file parsing solutions from open source RAG projects in the industry

深入 Dify 源码，洞察 Dify RAG 默认机制

Dive into the Dify source code and gain insight into the Dify RAG default mechanism

深入 Dify 源码，定位知识库检索的大模型调用异常

Go deep into the Dify source code and locate large model call anomalies in knowledge base retrieval

基于开源项目二次开发可行方案

The solution for secondary development based on open source projects

Dify 与 Xinference 最佳组合 GPU 环境部署全流程

The whole process of deploying the combination of Dify and Xinference in GPU environment

来自工业界的开源知识库 RAG 项目最全细节对比

The most complete comparison of open source RAG projects from the industry

官方技术文档查询大模型：快速搭建与实践记录

Official technical document query large model: rapid construction and practice records

html

深入源码，洞察迭代 8 年的 html 文本转换库

Dive into the source code and gain insights into the html text conversion library that has been iterating for 8 years

深入源码，洞察迭代 14 年的高质量 html 提取方案

Go deep into the source code and gain insight into the high-quality HTML extraction solution that has been iterated for 14 years

llm

AIOps RAG 比赛获奖项目 EasyRAG 深度解读

In-depth interpretation of AIOps RAG competition winning project EasyRAG

RAG 分块长距离信息缺失，Late Chunking 值得试试

Long distance contextual information is missing in RAG, you can try Late Chunking

No Code 的大模型应用跟踪与批量测试方案

No Code's large model application tracking and batch testing solution

你的大模型应用表现真的好吗？借助 Dify + Langfuse 一探究竟

Is your large model application really performing well? Find out with Dify + Langfuse

给大模型加上记忆，深入探索 Mem0 项目

Adding memory to large models, exploring the Mem0 project

langfuse

No Code 的大模型应用跟踪与批量测试方案

No Code's large model application tracking and batch testing solution

你的大模型应用表现真的好吗？借助 Dify + Langfuse 一探究竟

Is your large model application really performing well? Find out with Dify + Langfuse

RAG

AIOps RAG 比赛获奖项目 EasyRAG 深度解读

In-depth interpretation of AIOps RAG competition winning project EasyRAG

RAG 分块长距离信息缺失，Late Chunking 值得试试

Long distance contextual information is missing in RAG, you can try Late Chunking