Jungle-liu

Results 5 issues of Jungle-liu

请问训练出来一个多任务学习文本分类模型后,测试集里面的数据需不需要先确定是那个任务的数据在进行测试;还是一个完全无标签的数据,通过多任务学习模型会先给它确定在那个任务,再进行分类?

trainnum 4 4 4 79993 train: 4 vali: 4 test: 79993 AllTexts: 80001 text-ent: : 48366it [00:01, 28899.68it/s]

我想用wiki中文语料库来训练词向量,可以做到吗?那个word_strokes.txt,请问是怎么得到的呢?

Before updating the code, you can run the code normally, but after updating the code, what is the reason for the lack of real memory?