Skip to content

A user-friendly open-source project for recommendation systems.

License

Notifications You must be signed in to change notification settings

Iamctb/EasyDeepRecommand

Repository files navigation

EasyDeepRecommand

一个通俗易懂的开源推荐系统(A user-friendly open-source project for recommendation systems).

本项目将使用结合:代码、数据流转图、博客、模型发展史 等多个方面通俗易懂地讲解经典推荐模型,让读者通过一个项目了解推荐系统概况!

Dataset

Name Preprocess_url Download Progress
Criteo criteo_preprocess.py: 预处理源代码 Download_URL Done
预处理说明

Model_Zoo

No. Publication Model Blog Paper Version
1 DLRS'16 WideDeep 白话WideDeep Wide & Deep Learning for Recommender Systems, Google torch
2 ADKDD'17 DCN Deep & Cross Network for Ad Click Predictions, Google torch

Dependencies

本项目环境主要有:

  • python=3.8.20
  • pytorch=1.13.0

其余安装包可以使用下面命令安装:

pip install -r requirements.txt

Quick Start

以Criteo数据集和WideDeep举例:

Step1: 数据预处理

cd DataProcess/criteo
python criteo_preprocess.py

样本数据是使用的Criteo一万条数据作为示例,在执行命令过程中,需要注意 数据集的路径

Step2: 训练模型

data_config.json 中配置数据集路径;

model_config.json 中配置模型信息;

然后运行下面命令即可:

cd ModelZoo/WideDeep/WideDeep_torch
python train.py

最后

如果你觉得还不错的话,请帮忙点个star🌟吧,感谢感谢!!! If you think it's good, please help out with a star🌟, thank you !!!

About

A user-friendly open-source project for recommendation systems.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages