poprox_recommender.model.nrms#

Classes

NRMS(config)

NRMS network.

class poprox_recommender.model.nrms.NRMS(config)#

Bases: Module

NRMS network. Input 1 + K candidate news and a list of user clicked news, produce the click probability.

forward(candidate_news, clicked_news, clicked, mode='train')#

Defines the computation performed at every call.

Should be overridden by all subclasses.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.

get_news_vector(news)#

Parameters:: news (Tensor) – The encoded news content (currently just titles).
Returns:: (shape) batch_size, word_embedding_dim
Return type:: Tensor

get_user_vector(clicked_news_vector)#

Parameters:: clicked_news_vector – batch_size, num_clicked_news_a_user, word_embedding_dim
Returns:: (shape) batch_size, word_embedding_dim

get_prediction(news_vector, user_vector)#

Parameters:

news_vector – candidate_size, word_embedding_dim
user_vector – word_embedding_dim

Returns:

candidate_size

Return type:

click_probability

Modules

`news_encoder`
`user_encoder`