capreolus.extractor.pooled_bertpassage

Module Contents

Classes

PooledBertPassage

Extracts passages from the document to be later consumed by a BERT based model.

Attributes

logger

capreolus.extractor.pooled_bertpassage.logger[source]
class capreolus.extractor.pooled_bertpassage.PooledBertPassage(config=None, provide=None, share_dependency_objects=False, build=True)[source]

Bases: capreolus.extractor.bertpassage.BertPassage

Extracts passages from the document to be later consumed by a BERT based model. Different from BertPassage in the sense that all the passages from a document “stick together” during training - the resulting feature always have the shape (batch, num_passages, maxseqlen) - and this allows the reranker to pool over passages from the same document during training

module_name = pooledbertpassage[source]
dependencies[source]
config_spec[source]
create_tf_train_feature(self, sample)[source]

Returns a set of features from a doc. Of the num_passages passages that are present in a document, we use only a subset of it. params: sample - A dict where each entry has the shape [batch_size, num_passages, maxseqlen]

Returns a list of features. Each feature is a dict, and each value in the dict has the shape [batch_size, maxseqlen]. Yes, the output shape is different to the input shape because we sample from the passages.

create_tf_dev_feature(self, sample)[source]

Unlike the train feature, the dev set uses all passages. Both the input and the output are dicts with the shape [batch_size, num_passages, maxseqlen]

parse_tf_train_example(self, example_proto)[source]
parse_tf_dev_example(self, example_proto)[source]
id2vec(self, qid, posid, negid=None, label=None)[source]

See parent class for docstring