在＜百度框架师手把手教深度学习＞自然语言处理中的＂词向量word2Vec＂一课中．

定义SkipGram网络的时候

class SkipGram(fluid.dygraph.Layer):
    def __init__(self, vocab_size, embedding_size, init_scale=0.1):
        super(SkipGram, self).__init__()
        self.vocab_size = vocab_size
        self.embedding_size = embedding_size



        self.embedding = Embedding(
            size=[self.vocab_size, self.embedding_size],
            dtype='float32',
            param_attr=fluid.ParamAttr(
                name='embedding_para',
                initializer=fluid.initializer.UniformInitializer(
                    low=-0.5/embedding_size, high=0.5/embedding_size)))


        self.embedding_out = Embedding(
            size=[self.vocab_size, self.embedding_size],
            dtype='float32',
            param_attr=fluid.ParamAttr(
                name='embedding_out_para',
                initializer=fluid.initializer.UniformInitializer(
                    low=-0.5/embedding_size, high=0.5/embedding_size)))

同时也在前向计算中有

    def forward(self, center_words, target_words, label):
        center_words_emb = self.embedding(center_words)
        target_words_emb = self.embedding_out(target_words)

后面还有代码，但是我想要问的是，

forward()方法中    self.embedding(center_words)　　是在SkipGram中定义的么？为什么会传入参数center_words,SkipGram中的embedding（）好像没有参数的接收？

感谢阅读，并诚挚的希望收到您的回复