recommender-system

I am finding a product which can replace Elasticsearch.

作者您好，關於 AttRec 這份模型我有些疑問，就是在預測時會使用到 Next positive item 的 embedding，照理來說預測時是不能夠把真正的 Next positive item 輸入給模型，這樣等於是把正確答案丟給模型，請問真正在預測時，要如何得到 pos_scores 和 neg_scores 呢 ?

以下是程式碼中使用到 pos_embed 時，會讓我產生的疑問：

 # combine
 pos_scores = self.w * tf.reduce_sum(tf.multiply(short_interest, pos_embed), axis=-1, keepdims=True) \+ 
                        (1 - self.w) * tf.reduce_sum(pos_long_intere

Is your feature request related to a problem? Please describe.
There are cases, when it might be useful to know how much points are there in index, which satisfies some filtering condition.
Currently, we can only estimate a total amount of points.

Describe the solution you'd like
Create an API which returns amount of points (estimated or precise, depending on flag) for a given filter

Came up in jfkirk/tensorrec#31

It would be nice to have an arg to re-order the batches every epoch while fitting.

• Add shuffle_batches arg to fit() and fit_partial() that shuffles the batch order every epoch if True
• Add tests for the arg