Python Vocab.build_embedding_matrix方法代码示例

本文整理汇总了Python中utils.Vocab.build_embedding_matrix方法的典型用法代码示例。如果您正苦于以下问题：Python Vocab.build_embedding_matrix方法的具体用法？Python Vocab.build_embedding_matrix怎么用？Python Vocab.build_embedding_matrix使用的例子？那么恭喜您, 这里精选的方法代码示例或许可以为您提供帮助。您也可以进一步了解该方法所在类utils.Vocab的用法示例。

在下文中一共展示了Vocab.build_embedding_matrix方法的2个代码示例，这些例子默认根据受欢迎程度排序。您可以为喜欢或者感觉有用的代码点赞，您的评价将有助于系统推荐出更棒的Python代码示例。

示例1: Model

# 需要导入模块: from utils import Vocab [as 别名]
# 或者: from utils.Vocab import build_embedding_matrix [as 别名]
class Model():


    def __init__(self, config):
        self.config = config
        self.load_data(debug=False)
        self.build_model()


    def load_vocab(self,debug):
        self.vocab = Vocab()
        if debug:
            self.vocab.construct(get_words_dataset('dev'))
        else:
            self.vocab.construct(get_words_dataset('train'))
        self.vocab.build_embedding_matrix(self.config.word_embed_size)
        self.embedding_matrix = self.vocab.embedding_matrix


    def load_data(self, debug=False):
        """
            Loads starter word-vectors and train/dev/test data.
        """
        self.load_vocab(debug)
        config = self.config

        if debug:
            # Load the training set
            train_data = list(get_sentences_dataset(self.vocab,
                config.sent_len, 'dev', 'post'))
            ( self.sent1_train, self.sent2_train, self.len1_train,
                self.len2_train, self.y_train ) = zip(*train_data)
            self.sent1_train, self.sent2_train = np.vstack(self.sent1_train), np.vstack(self.sent2_train)
            self.len1_train, self.len2_train = ( np.array(self.len1_train),
                np.array(self.len2_train) )
            self.y_train = np.array(self.y_train)
            print('# training examples: %d' %len(self.y_train))

            # Load the validation set
            dev_data = list(get_sentences_dataset(self.vocab, config.sent_len,
                'test', 'post'))
            ( self.sent1_dev, self.sent2_dev, self.len1_dev,
                self.len2_dev, self.y_dev ) = zip(*dev_data)
            self.sent1_dev, self.sent2_dev = np.vstack(self.sent1_dev), np.vstack(self.sent2_dev)
            self.len1_dev, self.len2_dev = ( np.array(self.len1_dev),
                np.array(self.len2_dev) )
            self.y_dev = np.array(self.y_dev)
            print('# dev examples: %d' %len(self.y_dev))

            # Load the test set
            test_data = list(get_sentences_dataset(self.vocab, config.sent_len,
                'test', 'post'))
            ( self.sent1_test, self.sent2_test, self.len1_test,
                self.len2_test, self.y_test ) = zip(*test_data)
            self.sent1_test, self.sent2_test = np.vstack(self.sent1_test), np.vstack(self.sent2_test)
            self.len1_test, self.len2_test = ( np.array(self.len1_test),
                np.array(self.len2_test) )
            self.y_test = np.array(self.y_test)
            print('# test examples: %d' %len(self.y_test))
        else:
            # Load the training set
            train_data = list(get_sentences_dataset(self.vocab,
                config.sent_len, 'train', 'post'))
            ( self.sent1_train, self.sent2_train, self.len1_train,
                self.len2_train, self.y_train ) = zip(*train_data)
            self.sent1_train, self.sent2_train = np.vstack(self.sent1_train), np.vstack(self.sent2_train)
            self.len1_train, self.len2_train = ( np.array(self.len1_train),
                np.array(self.len2_train) )
            self.y_train = np.array(self.y_train)
            print('# training examples: %d' %len(self.y_train))

            # Load the validation set
            dev_data = list(get_sentences_dataset(self.vocab, config.sent_len,
                'dev', 'post'))
            ( self.sent1_dev, self.sent2_dev, self.len1_dev,
                self.len2_dev, self.y_dev ) = zip(*dev_data)
            self.sent1_dev, self.sent2_dev = np.vstack(self.sent1_dev), np.vstack(self.sent2_dev)
            self.len1_dev, self.len2_dev = ( np.array(self.len1_dev),
                np.array(self.len2_dev) )
            self.y_dev = np.array(self.y_dev)
            print('# dev examples: %d' %len(self.y_dev))

            # Load the test set
            test_data = list(get_sentences_dataset(self.vocab, config.sent_len,
                'test', 'post'))
            ( self.sent1_test, self.sent2_test, self.len1_test,
                self.len2_test, self.y_test ) = zip(*test_data)
            self.sent1_test, self.sent2_test = np.vstack(self.sent1_test), np.vstack(self.sent2_test)
            self.len1_test, self.len2_test = ( np.array(self.len1_test),
                np.array(self.len2_test) )
            self.y_test = np.array(self.y_test)
            print('# test examples: %d' %len(self.y_test))

            print('min len: ', np.min(self.len2_train))


    def build_model(self):
        config = self.config
        k = config.sentence_embed_size
        L = config.sent_len
#.........这里部分代码省略.........

开发者ID:cdelichy92，项目名称:DeepLearning-NLP-Project，代码行数:103，代码来源:shallow_attention_fusion_lstmn.py

示例2: Model

# 需要导入模块: from utils import Vocab [as 别名]
# 或者: from utils.Vocab import build_embedding_matrix [as 别名]
class Model():


    def __init__(self, config):
        self.config = config
        self.load_data()
        self.build_model()


    def load_vocab(self,debug):
        self.vocab = Vocab()
        if debug:
            self.vocab.construct(get_words_dataset('dev'))
        else:
            self.vocab.construct(get_words_dataset('train'))
        self.vocab.build_embedding_matrix(self.config.word_embed_size)
        self.embedding_matrix = self.vocab.embedding_matrix


    def load_data(self, debug=False):
        """
            Loads starter word-vectors and train/dev/test data.
        """
        self.load_vocab(debug)
        config = self.config

        if debug:
            # Load the training set
            train_data = list(get_sentences_dataset(self.vocab,
                config.sent_len, 'dev', 'post'))
            ( self.sent1_train, self.sent2_train, self.len1_train,
                self.len2_train, self.y_train ) = zip(*train_data)
            self.sent1_train, self.sent2_train = np.vstack(self.sent1_train), np.vstack(self.sent2_train)
            self.len1_train, self.len2_train = ( np.array(self.len1_train),
                np.array(self.len2_train) )
            self.y_train = np.array(self.y_train)
            print('# training examples: %d' %len(self.y_train))

            # Load the validation set
            dev_data = list(get_sentences_dataset(self.vocab, config.sent_len,
                'test', 'post'))
            ( self.sent1_dev, self.sent2_dev, self.len1_dev,
                self.len2_dev, self.y_dev ) = zip(*dev_data)
            self.sent1_dev, self.sent2_dev = np.vstack(self.sent1_dev), np.vstack(self.sent2_dev)
            self.len1_dev, self.len2_dev = ( np.array(self.len1_dev),
                np.array(self.len2_dev) )
            self.y_dev = np.array(self.y_dev)
            print('# dev examples: %d' %len(self.y_dev))

            # Load the test set
            test_data = list(get_sentences_dataset(self.vocab, config.sent_len,
                'test', 'post'))
            ( self.sent1_test, self.sent2_test, self.len1_test,
                self.len2_test, self.y_test ) = zip(*test_data)
            self.sent1_test, self.sent2_test = np.vstack(self.sent1_test), np.vstack(self.sent2_test)
            self.len1_test, self.len2_test = ( np.array(self.len1_test),
                np.array(self.len2_test) )
            self.y_test = np.array(self.y_test)
            print('# test examples: %d' %len(self.y_test))
        else:
            # Load the training set
            train_data = list(get_sentences_dataset(self.vocab,
                config.sent_len, 'train', 'post'))
            ( self.sent1_train, self.sent2_train, self.len1_train,
                self.len2_train, self.y_train ) = zip(*train_data)
            self.sent1_train, self.sent2_train = np.vstack(self.sent1_train), np.vstack(self.sent2_train)
            self.len1_train, self.len2_train = ( np.array(self.len1_train),
                np.array(self.len2_train) )
            self.y_train = np.array(self.y_train)
            print('# training examples: %d' %len(self.y_train))

            # Load the validation set
            dev_data = list(get_sentences_dataset(self.vocab, config.sent_len,
                'dev', 'post'))
            ( self.sent1_dev, self.sent2_dev, self.len1_dev,
                self.len2_dev, self.y_dev ) = zip(*dev_data)
            self.sent1_dev, self.sent2_dev = np.vstack(self.sent1_dev), np.vstack(self.sent2_dev)
            self.len1_dev, self.len2_dev = ( np.array(self.len1_dev),
                np.array(self.len2_dev) )
            self.y_dev = np.array(self.y_dev)
            print('# dev examples: %d' %len(self.y_dev))

            # Load the test set
            test_data = list(get_sentences_dataset(self.vocab, config.sent_len,
                'test', 'post'))
            ( self.sent1_test, self.sent2_test, self.len1_test,
                self.len2_test, self.y_test ) = zip(*test_data)
            self.sent1_test, self.sent2_test = np.vstack(self.sent1_test), np.vstack(self.sent2_test)
            self.len1_test, self.len2_test = ( np.array(self.len1_test),
                np.array(self.len2_test) )
            self.y_test = np.array(self.y_test)
            print('# test examples: %d' %len(self.y_test))

            print('min len: ', np.min(self.len2_train))


    def build_model(self):
        config = self.config
        k = config.sentence_embed_size
        L = config.sent_len
#.........这里部分代码省略.........

开发者ID:cdelichy92，项目名称:DeepLearning-NLP-Project，代码行数:103，代码来源:word_attention_model.py

注：本文中的utils.Vocab.build_embedding_matrix方法示例由纯净天空整理自Github/MSDocs等开源代码及文档管理平台，相关代码片段筛选自各路编程大神贡献的开源项目，源码版权归原作者所有，传播和使用请参考对应项目的License；未经允许，请勿转载。