o i @sddlmZmZmZddlmZmZmZmZm Z m Z m Z m Z m Z mZmZmZmZddlmZmZddlmZddlmZddlmZdd eeeffd d Zd ed eeeffddZdeeeeefdeeefdeeefd eeeeefeffddZd S))ListTuplecast) LinearLogisticMaxoutModelchain concatenateglorot_uniform_init list2ragged reduce_first reduce_last reduce_max reduce_mean with_getitem)Floats2dRagged)Doc)registry) extract_spansNreturncCstt||tdtS)zrAn output layer for multi-label classification. It uses a linear layer followed by a logistic activation. )nOnIinit_W)r rr r)rrrK/home/ubuntu/.local/lib/python3.10/site-packages/spacy/ml/models/spancat.pybuild_linear_logisticsr hidden_sizecCsBttttttftttttfttt t |dddS)zReduce sequences by concatenating their mean and max pooled vectors, and then combine the concatenated vectors with a hidden layer. Tg)r normalizedropout) r r rrrrrr rrr)r rrrbuild_mean_max_reducer s r#tok2vecreducerscorerc Csztttttttftttfftdt|tttttft t ||}| d|| d|| d||S)aBuild a span categorizer model, given a token-to-vector model, a reducer model to map the sequence of vectors for each span down to a single vector, and a scorer model to map the vectors to probabilities. tok2vec (Model[List[Doc], List[Floats2d]]): The tok2vec model. reducer (Model[Ragged, Floats2d]): The reducer model. scorer (Model[Floats2d, Floats2d]): The scorer model. rr$r%r&) r rrrrrrrrr rset_ref)r$r%r&modelrrrbuild_spancat_model/s   r))NN)typingrrr thinc.apirrrrr r r r r rrrr thinc.typesrrtokensrutilrrrintr#r)rrrrs"<