Bert自然语言处理(NLP)

feature based和fine tuning

2019-07-08  本文已影响3人  Leahlijuan

相同点

都是借助别人已有的NLP模型完成自己的任务

不同点

As shown in figure 2 of {1}, in the fine-tuning strategy all weights are changed when training on the new task (except for the weights of the last layers for the original task), whereas in the feature extraction strategy only the weights of the newly added last layers change during the training phase:

image.png
feature-baed and fine tuning

feature-based

只变化了最后一层的参数。
通常feature-based方法包括两步:

EMLO是这种方法的典型

fine-tuning

除了最后一层,所有的参数都变化了。
Fine-tuning方式是指在已经训练好的语言模型的基础上,加入少量的task-specific parameters, 例如对于分类问题在语言模型基础上加一层softmax网络,然后在新的语料上重新训练来进行fine-tune。

参考资料知乎fine-tuning

上一篇下一篇

猜你喜欢

热点阅读