About recommender system

[yanchaoguo]

请问有没有推荐系统中召回或排序的例子以及文本分类的例子

[frankfliu]

You may want to take a look this example: https://github.com/deepjavalibrary/djl-demo/tree/master/apache-beam/ctr-prediction

You should be able to find many regression model for XGBoost, DJL support XGBoost now, you can use XGBoost train your model and use DJL to run XGBoost model

We also support fastText/blazingText, you can find text classification example here: https://github.com/deepjavalibrary/djl/blob/master/extensions/fasttext/src/test/java/ai/djl/fasttext/CookingStackExchangeTest.java

[yanchaoguo]

You may want to take a look this example: https://github.com/deepjavalibrary/djl-demo/tree/master/apache-beam/ctr-prediction

You should be able to find many regression model for XGBoost, DJL support XGBoost now, you can use XGBoost train your model and use DJL to run XGBoost model

We also support fastText/blazingText, you can find text classification example here: https://github.com/deepjavalibrary/djl/blob/master/extensions/fasttext/src/test/java/ai/djl/fasttext/CookingStackExchangeTest.java

Ctr这个例子里说是deepfm 可在inference阶段 没看到modelzoo通过Criteria加载deepfm模型啊 莫非默认就是吗

[frankfliu]

In the example, you need specify system properties: -Dai.djl.repository.zoo.location=./deepfm

This will tell DJL to scan ./deepfm folder to find models.

[yanchaoguo]

In the example, you need specify system properties: -Dai.djl.repository.zoo.location=./deepfm

This will tell DJL to scan ./deepfm folder to find models.

通过Criteria设置modelpath也可以加载这个模型吧，还有拿到模型网络结构后如何使用djl处理自己的数据集 得到一个不错的模型呢

[treper]

这方面的例子有点少。。