[Adapt] [seminar] An Introduction on demonstration learning

[Brandon]

Hi Adapters,


Pretraining + finetuning paradigm has brought large improvements for many natural language processing tasks. However, if there is not enough training data, the performance of the model will be poor. Prompt-based learning is based on language models that model the probability of text directly. By defining a new prompting function the model is able to perform few-shot or even zero-shot learning, adapting to new scenarios with few or no labeled data. Demonstration is an important technology in prompt-based learning. The idea of demonstration is to give the model some demos when doing inference. I will present an brief introduction on demonstration learning.


Hope you enjoy it!


Time: Wed 4:00 pm Venue: SEIEE 3-414Best Regards,
Bran
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://cs.sjtu.edu.cn/pipermail/adapt/attachments/20211215/8214d82b/attachment.htm>