加载预训练词向量 trainable=False
Created by: gongel
word_emb_pretrained = fluid.layers.embedding( input=word, size=[word_dict_len, 32], dtype='float32', is_sparse=IS_SPARSE, param_attr=fluid.ParamAttr( learning_rate=0, name="word_emb_pretrained", initializer=uniform_initializer, trainable=False))
通过以上方式加载了预训练的词向量,并保证在下游任务中不被训练,请问怎么查看预训练词向量在训练中是否改变?谢谢