Git Product home page Git Product logo

adt's Introduction

Adaptive Disentangled Transformer

Overall description

This is the repository of the KDD 2023 paper "Adaptive Disentangled Trasnformer for Sequential Recommendation"

This repository reimplements the backbone model from SASRecBert4Rec and STOSA.

Requirements

The following libraries and versions are used in the experiments, these requirements are not mandatory and are only for reference.

numpy = 1.22.4
Python = 3.8.10
Pytorch = 1.11.0
scipy = 1.9.3
tqdm = 4.61.2

Usage

Search

For each backbone, you can use this command to search the best lambdas for training the model. You can go the source file for more details.

python evolution.py --dataset xxx

Retrain

After searching process, you will get the best candidate in /res and you can copy the candidate vector and your search space to candidates_to_lambdas.py and run it to get the best lambdas, or copy the value of key rec and ind to get the best lambdas. And then copy the lambdas to function get_lambdas in utils.py and retrain the model to get the final score.

For convenience, we have recorded part of the lambdas used in our experiments in utils.py and hyperparameters in /templates as an example.

python main.py --dataset xxx

Data

Part of the datasets used for SASRec backbone and Bert4Rec backbone are uploaded to the repository, and ml-20m can download from this link. For STOSA backbone, you can download the 5-core dataset from Amazon Dataset or go to the original repository for the details.

adt's People

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

Watchers

 avatar  avatar

Forkers

asachanmanchoi

adt's Issues

decoder部分输入没有看懂

师兄你好,你的decoder输入的是什么啊?论文里面说它的输入是the first 𝑇 − 1 time steps of the input sequence of the encoder,所以encoder和decoder的输入有什么区别吗?为什么要用这个作为输入啊,一般的transformer模型不都是用encoder的输出作为decoder的输入吗?

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.