Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/bruceunx/deepfm
simple deepfm
https://github.com/bruceunx/deepfm
Last synced: 6 days ago
JSON representation
simple deepfm
- Host: GitHub
- URL: https://github.com/bruceunx/deepfm
- Owner: bruceunx
- Created: 2022-06-22T03:18:05.000Z (over 2 years ago)
- Default Branch: master
- Last Pushed: 2022-06-22T03:18:21.000Z (over 2 years ago)
- Last Synced: 2023-10-15T23:24:32.016Z (about 1 year ago)
- Language: Python
- Size: 105 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
## TIPS
- 程序开发采用python, pytorch
- models.py 模型定义代码文件
- dataset.py 数据dataset代码文件
- main.py 主程序,训练model
## 简介
### models.py
- Fm `fm 部分参考文献和torchrec`
- Dense `线性层构建的deep部分`
- Deepfm `embedding层 -> feature embedding >(dense, fm) -> out`### main.py
- 定义sparse feats 字典,用来定义embedding层的维度
- 目前文件中dense feats 为空, userID数据庞大,引入embedding层,可能使参数非常大,不方便训练
- 定义 hidden units 为 Dense 参数
- 定义auc, logloss函数
- 考虑数据量太大,采用chunsize迭代dataframe处理,但是要考虑embedding层的维度
- 将数据拆分train, val, 训练模型