Skip to main content

DeepMatch-Torch is a PyTorch Version deep matching model library for recommendations & advertising. It's easy to train models and to export representation vectors for user and item which can be used for ANN search.

Project description

DeepMatch-Torch

Python Version PyTorch Version PyTorch-Lightning PyPI

DeepMatch-Torch is a PyTorch version of DeepMatch.

DeepMatch-Torch is a PyTorch Version deep matching model library for recommendations & advertising. It's easy to train models and to export representation vectors for user and item which can be used for ANN search. You can use any complex model with model.fit() and model.predict(). And you can keep same habit of using DeepMatch.

Let's Get Started! or Run examples !

1. Install

  1. Install deepmatch-torch by pip
pip install deepmatch-torch
  1. Install deepmatch-torch through source code
git clone https://github.com/bbruceyuan/DeepMatch-Torch.git
cd DeepMatch-Torch
python setup.py install

2. Tutorial

You can run example by following steps.

cd examples
python run_youtubednn.py or run_fm_dssm.py

3. Models List

Model Paper
FM [ICDM 2010]Factorization Machines
DSSM [CIKM 2013]Deep Structured Semantic Models for Web Search using Clickthrough Data
YoutubeDNN [RecSys 2016]Deep Neural Networks for YouTube Recommendations
NCF [WWW 2017]Neural Collaborative Filtering
MIND [CIKM 2019]Multi-interest network with dynamic routing for recommendation at Tmall

TODO

  • simplify model config. now only support kwargs, but config is a elegant choice.
  • fix MIND only support CPU train bug.

Acknowledgments

Especially thanks to DeepMatch. This project relies highly on DeepMatch. Additionally, I used the PLBaseModel design from torchTS.

Thanks to this awesome projects.

中文 README

DeepMatch-Torch 是一个 PyTorch 版本的 DeepMatch

DeepMatch-Torch 是一个用于广告推荐的召回模型库,可以非常简单地训练模型和导致 useritemvertor 表示,你可以用这个 user/item 的表示进行 ANN 近似检索。在 DeepMatch-Torch 中,你可以保持和 DeepMatch 一样的习惯,通过 model.fit()model.predict() 进行模型的训练和预测。

更多的中文文档细节可以参见:中文文档

1. 安装

  • 步骤一: 安装 PyTorch, 按照官网指引
  • 步骤二:
    • 通过 pip 安装:pip install deepmatch-torch
    • 通过源码安装
git clone https://github.com/bbruceyuan/DeepMatch-Torch.git
cd DeepMatch-Torch
python setup.py install

2. 快速上手

进入 example 目录查看相关代码,使用方式和 DeepMatch 几乎保持一致。

cd examples
python run_youtubednn.py or run_fm_dssm.py

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

deepmatch-torch-0.0.6.tar.gz (15.6 kB view details)

Uploaded Source

Built Distribution

deepmatch_torch-0.0.6-py3-none-any.whl (27.8 kB view details)

Uploaded Python 3

File details

Details for the file deepmatch-torch-0.0.6.tar.gz.

File metadata

  • Download URL: deepmatch-torch-0.0.6.tar.gz
  • Upload date:
  • Size: 15.6 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.8.0 pkginfo/1.8.2 readme-renderer/32.0 requests/2.26.0 requests-toolbelt/0.9.1 urllib3/1.26.7 tqdm/4.62.3 importlib-metadata/4.8.2 keyring/23.5.0 rfc3986/2.0.0 colorama/0.4.4 CPython/3.9.7

File hashes

Hashes for deepmatch-torch-0.0.6.tar.gz
Algorithm Hash digest
SHA256 b3185141e53250b1ab30794350c4bb5256b7505287c153977c531b44bdc37205
MD5 2ea55757341a9cc4c999f9246b65476f
BLAKE2b-256 32bd9367974a16d8adc19055ea76747e245f8cf3713f9dfbed55f0c417b4e306

See more details on using hashes here.

File details

Details for the file deepmatch_torch-0.0.6-py3-none-any.whl.

File metadata

  • Download URL: deepmatch_torch-0.0.6-py3-none-any.whl
  • Upload date:
  • Size: 27.8 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.8.0 pkginfo/1.8.2 readme-renderer/32.0 requests/2.26.0 requests-toolbelt/0.9.1 urllib3/1.26.7 tqdm/4.62.3 importlib-metadata/4.8.2 keyring/23.5.0 rfc3986/2.0.0 colorama/0.4.4 CPython/3.9.7

File hashes

Hashes for deepmatch_torch-0.0.6-py3-none-any.whl
Algorithm Hash digest
SHA256 7c8292e5a8b64e6a3736a31d16a6c22af99be164093331cd4f488fbc1d078261
MD5 d4675ef04f8e68aaf8a400b2abd0edfe
BLAKE2b-256 6c265dcf81042a4995ccdd428f186a5a6fba753ae7ff27fb96dc113fb9f6ca68

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page